The following Essay is part of a mini-symposium on the 2020 Census and its broader implications for our nation and democracy. Both essays rely on the perspectives of scholar-practitioners to interrogate the design of the enumeration process and outline its potential ramifications for the distribution of federal resources and political power in an evolving United States.

*  *  *

The constitutionally mandated decennial enumeration of the U.S. population is indispensable to the equitable distribution of political and economic resources. As we approach the 2020 Census, however, several factors converge that both undermine how we count change in commu­nities of color and conflict with shifting demographics and power dynamics, making accurate accounting especially urgent. Among these, perhaps most notable is the threatened inclusion of a citizenship question on the Census short form. By 2045, the population of the United States is on pace to comprise a majority-minority plurality in which no single racial or ethnic group constitutes a majority but people of color, collectively, outnumber whites. Accurately counting this demo­graphic change will help determine the appropriate allocation of the resources most needed to serve an evolving American populace. This Essay considers the historical context of undercounting people of color in the United States, the mounting stakes of their continued numerical diminution, and certain policy and pragmatic measures to disrupt the effects of a compromised Census. Premised on a theory of representa­tional equality in which all our country’s residents are counted and served as constituents, this Essay provides a comprehensive assessment of the central role the Census plays in our democracy in an effort to ensure that we do not discount the diversity that makes this country the greatest and most promising democratic experiment to date.


    1. Rights Enforcement
    2. Allocation of Federal Funds
    3. Reapportionment and Redistricting
    4. Changing Demographics
    1. The Citizenship Question
    2. Prison-based Gerrymandering
    3. Displacement, Downturn, and Natural Disaster
    4. Digital Divide
    1. Use Interest Convergence to Broker Power on Behalf of the Emerging Majority
    2. Leverage Voter Turnout Efforts for the 2018 Midterm Elections to Create a Census Mobilization Apparatus
    3. Pass Omnibus Legislation Creating Nonpartisan Redistricting Commissions, Mandating Sampling, and Implementing a “Usual Residence” Rule for Incarcerated Persons
    4. Engage Private Sector Support to Enhance Accurate Census Enumeration



An accurate, periodic accounting of the U.S. population is indispen­sable to the equitable distribution of political and economic resources. These resources range from the allocation of nearly $700 bil­lion in federal funds to the constitutionally mandated apportionment of con­gressional representation. 1 Pursuant to Article I, Section 2 of the U.S. Constitution, “[r]epresentatives . . . shall be apportioned among the several States . . . according to their respective Numbers.” U.S. Const. art. I, § 2, cl. 3. The decennial census is the process through which the federal government audits the U.S. population for these and myriad other purposes. As we approach the 2020 Census, however, several factors converge that both undermine how we count change in communi­ties of color and conflict with shifting demographics and power dynamics, making an accurate accounting especially urgent.

By 2045, the population of the United States is on pace to comprise a majority-minority plurality in which no single racial or ethnic group constitutes a majority but people of color, collectively, outnumber whites. 2 See William H. Frey, The US Will Become ‘Minority White’ in 2045, Census Projects, Brookings Inst. (Mar. 14, 2018), [] (citing 2018 U.S. Census statistics that suggest the country will be “minority white” in 2045). Census statistics predict that by 2045, “whites will comprise 49.7 percent of the population in contrast to 24.6 percent for Hispanics, 13.1 percent for blacks, 7.9 percent for Asians, and 3.8 percent for multiracial populations.” Id. This would mark the first time in the history of the U.S. Census that white residents would not constitute a numerical majority of the total pop­ula­tion. 3 William H. Frey, US White Population Declines and Generation ‘Z-Plus’ Is Minority White, Census Shows, Brookings Inst. (June 22, 2018), [] (“[F]or the first time since the Census Bureau has released these annual statistics, they show an absolute decline in the nation’s white non-Hispanic population . . . . This means we are on the cusp of seeing the first minority white generation . . . .”). Nonwhite persons already constitute 39% of the population, and that figure is forecast to continue to trend upward as a result of high birth rates among communities of color and rising death rates among the aging population of whites. 4 See Jonathan Vespa, David M. Armstrong & Lauren Medina, U.S. Census Bureau, Demographic Turning Points for the United States: Population Projections for 2020 to 2060, at 3 (2018),
demo/P25_1144.pdf [] (“The non-Hispanic White population is projected to shrink over coming decades, from 199 million in 2020 to 179 million people in 2060—even as the U.S. population continues to grow. Their decline is driven by falling birth rates and rising number of deaths . . . as the non-Hispanic White population ages.”).
As one commentator has described it, “the demographic revolution is now irreversible. . . . [T]he country is . . . inexorably on a path to a new multiracial reality.” 5 Steve Phillips, Opinion, Trump Wants to Make America White Again, N.Y. Times (Feb. 15, 2018),
white-again.html (on file with the Columbia Law Review).

The U.S. government must accurately count its electorate if it is to adequately prepare for the consequences of this impending demo­graphic facelift. Not only is the total population of the country increas­ing, 6 See Vespa et al., supra note 4, at 1 (“Despite slowing population growth, particu­larly after 2030, the U.S. population is still expected to grow by 78 million people by 2060, cross­ing the 400-million threshold in 2058.”). creating a greater overall demand on federal resources, but the number of people of color who rely disproportionately on certain federal resources 7 For example, of the seventy million Medicaid beneficiaries in 2016, 43% were white, 18% were black, and 30% were Latinx. Arthur Delaney & Ariel Edwards-Levy, Americans Are Mistaken About Who Gets Welfare, HuffPost (Feb. 5, 2018),
entry/5a7880cde4b0d3df1d13f60b []. Of the forty-three mil­lion recipients of food stamps under the Supplemental Nutrition Assistance Program that same year, 36.2% were white, 25.6% black, and 17.2% Latinx. Id. Similarly, of the 2.7 mil­lion people served by the Temporary Assistance for Needy Families program in 2016, 27.6% were white, 29.1% black, and 36.9% Latinx. Id. During this same time period, whites, blacks, and Latinx persons comprised 61%, 12%, and 18% of the population, respectively. Population Distribution by Race/Ethnicity, State Health Facts, 2017, Henry J. Kaiser Family Found., [https://] (last visited Jan. 28, 2019). It is important to note, however, that contrary to public perception, in raw numbers, whites are the largest consumers of public resources. See Delaney & Edwards-Levy, supra.
as a result of systemic discrimination and immigration status is also increasing. Accurately counting that change will help determine the appropriate allocation of the resources most needed to serve an evolving American populace. However, certain policy prescriptions stand in the way. Concerns that the country is becoming a majority-minority nation already appear to be influencing policy preferences. Nowhere, perhaps, are these influences more evident than in the policy choices that shape the U.S. Census—the very tool designed to enumerate the population and quantify the country’s changing demographics.

For the first time in seventy years, spanning both Democratic and Republican administrations, the U.S. Secretary of Commerce champi­oned and approved the inclusion of a citizenship question on the Census short form. 8 Susan McFarland, For the First Time in 70 Years, 2020 Census to Ask About Citizenship, UPI (Mar. 27, 2018), []. This change, set to be implemented in the 2020 Census, has been successfully challenged in three federal district courts. 9 In the consolidated cases of New York v. U.S. Department of Commerce and New York Immigration Coalition v. U.S. Department of Commerce, Judge Jesse M. Furman held that U.S. Secretary of Commerce Wilbur Ross violated the Administrative Procedure Act (APA) “in multiple independent ways.” New York v. U.S. Dep’t of Commerce, 351 F. Supp. 3d 502, 516 (S.D.N.Y. 2019), cert. granted, 139 S. Ct. 953 (2019) (mem.). Specifically, the court concluded that the APA required Secretary Ross to collect data “through the acquisition and use of ‘administrative records’ instead of through ‘direct inquiries’ on a survey such as the census.” Id. (emphasis omitted). Relying solely on the administrative record—consist­ing of materials the defendants submitted as evidence of what Secretary Ross pur­port­edly relied on in his decisionmaking—the court further held that “Secretary Ross’s decision to add a citizenship question was ‘arbitrary and capricious’ on its own terms”: The decision was not supported by any evidence in the record, and the stated justification of enabling compliance with the Voting Rights Act was “pretextual.” Id. While the court stopped short of holding that Secretary Ross’s stated justifications were pretext for intentional discrim­ination in violation of the Equal Protection Clause, it did note that plaintiffs were unable to meet their burden to prove such a violation in part because the Supreme Court held that Secretary Ross could not be deposed. Id. at 517. Following that decision, a district court in California similarly found that including the citizenship question on the short form vio­lated the APA and the Enumeration Clause of the Constitution. See California v. Ross, 358 F. Supp. 3d 965, 1041, 1048 (N.D. Cal. 2019). The court also held that Secretary Ross’s “pur­ported purpose [for including the citizenship question] was a mere pretext . . . [that] did not supply the true basis of Secretary Ross’s decision.” Id. at 1040. Most recently, a judge in the District of Maryland also found that the citizenship question violates both the Enumeration Clause of the Constitution and the APA, holding that
[t]he unreasonableness of Defendants’ addition of a citizenship ques­tion to the Census is underscored by the lack of any genuine need for the citizenship question, the woefully definicient process that led to it, the mysterious and potentially improper political considerations that moti­vated the decision and the clear pretext offered to the public.
Kravitz v. U.S. Dep’t of Commerce, Nos. GJH-18-1041, GJH-18-1570, 2019 WL 1510449, at *52 (D. Md. Apr. 5, 2019).
The battle is not yet over, however. The U.S. Supreme Court has already granted certiorari in one case and is expected to rule on the legality of the citizenship question before the end of the Term. 10 U.S. Dep’t of Commerce v. New York, 139 S. Ct. 953 (2019) (mem.). Thus, there is an active moral and legal imperative to explore the poten­tial impact of including a citizenship question on the Census for the first time in nearly three generations. Further, should the Supreme Court per­mit inclusion of the citizenship question, other actions become increas­ingly necessary to ensure an accurate enumeration of the population, most notably of communities of color.

In addition to being arbitrary and capricious, the decision to include the citizenship question on the Census short form may have been ani­mated by xenophobia, political animus, and racism, rather than sound policy and principle. 11 See, e.g., Jonathan Blitzer, The Motives Behind the Trump Administration’s New Census Question on Citizenship, New Yorker (Mar. 29, 2018), (on file with the Columbia Law Review) (describing the inclusion of the citizen­ship question as not “just political but existential”). Professor Rick Hasen has remarked that “[t]he idea that this question is being added to the census to protect minority voting rights so that the lines can be redrawn in a fair way is just too absurd to be believed.” Id.; see also Eric Anthony Licas, Citizenship Question on U.S. Census ’20 ‘Politically Motivated’—Critics, Inquirer.Net (July 20, 2018), [] (reporting that a coalition of civil rights groups asserted that the citizenship question was politically motivated and intended to depress Census participa­tion in communities of color). Indeed, on July 31, 2018, Bloomberg News’s editorial board issued a scathing op-ed critiquing the decision to include a citizenship question on the Census, calling into question the stated motivation for the question and concluding that it is intended to produce an undercount in states with large populations of Democratic voters. See Editorial, Trump Prepares to Bungle the 2020 Census: Adding a Citizenship Question Would Skew Its Findings—Especially in Blue States, Bloomberg (July 31, 2018),
a-bad-push-from-trump-and-ross (on file with the Columbia Law Review).
As Professor Justin Levitt has written, 12                         See Justin Levitt, Citizenship and the Census, 119 Colum. L. Rev. 1355 (2019). and all three district courts have found, 13                  SeeKravitz, 2019 WL 1510449, at *52; California, 358 F. Supp. 3d at 1040; New York, 351 F. Supp. 3d at 516. Despite finding that Secretary Ross’s decision “was pretextual,” the court in New York held that the plaintiffs, hampered by the Supreme Court’s stay of the dis­trict court’s order authorizing Secretary Ross’s deposition, did not prove an equal protec­tion violation based on invidious discrimination. New York, 351 F. Supp. 3d at 517 (citing In re Dep’t of Commerce, 139 S. Ct. 566 (2018) (mem.); In re Dep’t of Commerce, 139 S. Ct. 16, 16–17 (2018) (mem.)). The Supreme Court’s stay had the effect of excluding the best evidence of Secretary Ross’s intent. Id. Secretary Ross’s proffered justifica­tion—that including a citizenship question would improve enforcement of the Voting Rights Act—is both dubious as a policy proposition and indefen­sible as a practical matter. 14 See Levitt, supra note 12, at 1374–86. For the entire lifespan of the Voting Rights Act, the Census short form has not included a citizenship question. Nonetheless, during this time, civil rights groups like the NAACP Legal Defense and Educational Fund (LDF), where I serve as Associate Director-Counsel, have successfully used Census data to defend and even expand minority voting rights. 15                LDF and other civil rights organizations have submitted amicus briefs to the Supreme Court specifically attesting to the fact that Section 2 enforcement would be inhibited, not enhanced, by inclusion of a citizenship question on the short form. See Brief of Amicus Curiae NAACP Legal Defense & Educational Fund, Inc. in Support of Respondents at *30–37, Dep’t of Commerce v. New York, No. 18-966 (U.S. filed Apr. 1, 2019), 2019 WL 1472070 (arguing that, based on extensive experience litigating Section 2 claims, the citizen­ship question is unnecessary to enforce the Voting Rights Act). Indeed, in numerous cases the Supreme Court has upheld the sufficiency of existing Census data in prosecuting Section 2 claims. Id. at *34. There are credible concerns that, rather than protecting their voting rights, the citizenship question will force racial and ethnic minority communities into the shadows and combine with other factors to cause a severe and potentially incalculable undercount of people of color. 16 See infra section III.A.

As a separate but related matter, last year the Census Bureau indi­cated its intention to continue using “Residence Criteria” that count the more than two million incarcerated people in the United States in their location of confinement rather than in their pre-incarceration residence where they would otherwise live as full members of society. 17 See Final 2020 Census Residence Criteria and Residence Situations, 83 Fed. Reg. 5525, 5526 (Feb. 8, 2018) (to be codified at 15 C.F.R. ch. 1) (“[T]he concept of ‘usual resi­dence’ . . . is defined by the Census Bureau as the place where a person lives and sleeps most of the time. This is not always the same as a person’s legal residence, voting resi­dence, or where they prefer to be counted.”). In light of severe racial and ethnic disparities between incarcerated persons and the general population, 18 According to the Prison Policy Initiative, using data as of 2014, 35.4% of inmates in state and federal prisons were black and 21.6% were Latinx. See Alison Walsh, The Criminal Justice System Is Riddled with Racial Disparities, Prison Policy Initiative (Aug. 15, 2016), []. These numbers are particularly startling when compared to the percentage of the total U.S. population that each group constitutes: 13.2% and 17.4%, respectively. Id. the Residence Criteria adversely affect a dispropor­tionate number of black and Latinx 19 The term “Latinx” is a non-gender-binary reference to the group of persons com­monly identified as Latino/a or Hispanic. The Census uses the terms “Hispanic, Latino, or Spanish origin” to identify Latinx persons. See Kenneth Prewitt, What Is Your Race? The Census and Our Flawed Efforts to Classify Americans 175 fig.3 (2013). It also permits peo­ple who identify by one of those terms to check a box to further identify themselves as “Mexican, Puerto Rican, Cuban, or other Hispanic origin” or to write in a description. See id. Respondents are then asked to identify themselves by race by checking a box for white, black, American Indian or Alaska Native, one of seven Asian nationalities, one of four Pacific Islander groups, or to identify themselves as “some other race,” accompanied by a write-in box in which they can provide a description. See id. at 176 fig.4. In 2010, approxi­mately twenty million Census respondents, most of whom identified as Latinx, checked “some other race,” see id. at 175–76, making it the third-most selected race category following white and black. Sowmiya Ashok, The Rise of the American ‘Others,’ Atlantic (Aug. 27, 2016),
497690/ [] (“‘Some Other Race,’ . . . was the third-largest cat­egory after ‘White’ and ‘Black’ in 2010, alarming officials, who are concerned that if nothing is done ahead of the 2020 census, this non-categorizable category of people could become the second-largest racial group in the United States.”); see also Prewitt, supra, at 175, 183–208 (proposing alternative methods of identifying persons by race or ethnicity).
communities by counting their permanent residents in other areas, a disproportionate number of which are predominantly white. 20 See Gregory Hooks et al., The Prison Industry: Carceral Expansion and Employment in U.S. Counties, 1969–1994, 85 Soc. Sci. Q. 37, 40 (2004) (“The rural and predominantly white areas in which most prisons are located increase their share of federal grants and political representation; the impoverished urban communities that [black and Latinx] pris­oners leave behind lose funds and representation.”). Prison-based gerrymandering, as this process is formally known, tends to favor rural, white communities, which house a disproportionate number of prisons, and harm residents outside of those areas, especially urban communities populated by people of color, which are denser and tend to yield the majority of incarcerated persons. 21 See Prison-Based Gerrymandering Reform, NAACP LDF (Feb. 16, 2018), https:// []. Prison-based gerrymandering is defined as the “practice whereby many states and local governments count incarcerated persons as residents of the areas where they are housed when election district lines are drawn,” which, in turn, “distorts our democratic process by artificially inflating the population count—and thus, the political influence—of the districts where prisons and jails are located,” resulting in the diminution of the voting power of everyone living outside of those districts. Id. For more on prison-based gerryman­dering and its interrelation with the Census, see infra section III.B.

Because they collectively result in an inaccurate population count, these and other Census-related policy choices frustrate accurate appor­tionment, equitable allocation of federal resources, and the ability of states and localities to redistrict in a manner that complies with state constitutions, the U.S. Constitution, and the Voting Rights Act. These policy choices also reflect a certain theory of representation in our representative democracy that forces a reckoning with Justice Ginsburg’s definition of “representational equality,” under which “all residents of the state are to be counted—and served—as constituents.” 22 Janai Nelson, Evenwel v. Abbott: Every Person (and Every Win) Counts, Hamilton & Griffin on Rights (Apr. 10, 2016), []; see also Evenwel v. Abbott, 136 S. Ct. 1120, 1128–32 (2016). Justice Ginsburg’s expansive and inclusive theory of representation, which the Supreme Court validated in Evenwel v. Abbott, 23 Evenwel, 136 S. Ct. at 1128–29, 1132. is at odds with the practical realities of where and how we have historically counted certain groups of people, especially people of color; 24 See infra Part I. our present practices concerning the counting of incarcerated persons; 25 See infra section III.B. and the govern­ment’s rabid drive to insert an overwhelmingly unpopular and untested citizenship question on the Census short form. 26 See Census Bureau Overwhelmed by Public Comments Opposing Citizenship Question, The Census Project (Jan. 24, 2019), [https://] (reporting that of the 147,831 comments that the Census Bureau received on summer 2018’s Federal Register Notice, approximately 138,000 concerned the citizenship question, 99% of which opposed the question).

This piece considers the historical context of undercounting people of color in the United States, the mounting stakes surrounding their continued numerical diminution, and certain policy and pragmatic measures to disrupt the effects of a compromised Census. Part I outlines the fraught history of (dis)counting communities of color that dates back to the founding of our democracy. That history reveals that the prerogative of counting people of color is most easily exercised in their favor if it does not result in creating political accountability to them. In other words, decisions about whether and where to enumerate people of color have historically been influenced by whether that enumeration would shift the extant inequitable allocation of political power and in what direction.

Part II discusses the most compelling reasons for an accurate population count, not only for communities of color but for our nation as a whole. From the allocation of funding to congressional apportion­ment, from redistricting to civil rights enforcement, accurate Census data form the numerical foundation to fulfill constitutional mandates, exe­cute the country’s strategic plan, and meet its projections. Census data are also used “for such varied purposes as computing federal grant-in-aid benefits, drafting of legislation, urban and regional planning, business planning, and academic and social studies.” 27     Baldrige v. Shapiro, 455 U.S. 345, 353 n.9 (1982). Understanding these stakes crystalizes the imperative for accuracy in enumeration. Part III examines a range of barriers to an accurate count. It also examines the critical consequences of an inaccurate enumeration that the inclusion of a citizenship question would likely produce, the practice of prison-based gerrymandering, and other factors that threaten to further distort the Census count in 2020. Finally, Part IV prescribes certain actions and advocacy to counteract the likely undercount of commu­nities of color and improve the chances of counting change in the American demographic—if not exactly, then as capably as possible.

I. The Census in Historical Context

Counting residents of color in the United States has been a con­tested exercise since the country’s Founding. During certain periods in our history, population enumeration served as an inventory of enslaved Africans as human chattel and chronicled in numbers the genocide 28 See Donald L. Fixco, When Native Americans Were Slaughtered in the Name of ‘Civilization,’ History (Mar. 2, 2018), [] (detailing “some of the most aggres­sive acts of genocide taken against indigenous Americans” as well some of the reasons why this genocide occurred). of this country’s indigenous population. 29 During the early decades of the Census through 1960, Census data revealed negligible growth in Native American populations due to high mortality rates and low birth rates. See Jeffrey S. Passel, The Growing American Indian Population, 1960–1990: Beyond Demography, in Changing Numbers, Changing Needs: American Indian Demography and Public Health 79, 79 (Gary D. Sandefur et al. eds., 1996) (“For decades through 1960, the American Indian population, as enumerated in U.S. censuses, grew little if at all. . . . [T]he average annual growth rate over the entire 70-year period was only 1.1 percent—a very low figure resulting from high fertility and very high mortality.” (footnote omitted)). The  Census  has  also  recorded  the  migration  of  freed  persons  seeking  to  fully  inhabit  their  citizenship, 30 See, e.g., The Great Migration, 1910 to 1970, U.S. Census Bureau (Sept. 13, 2012), [] (“The Great Migration generally refers to the massive internal migration of Blacks from the South to urban centers in other parts of the country. Between 1910 and 1970, an esti­mat­ed 6 mil­lion Blacks left the South [to seek opportunity in the Midwest and the North].”); see also Isabel Wilkerson, The Warmth of Other Suns: The Epic Story of America’s Great Migration 8–15 (2010) (describing the motivations for the Great Migration from 1915 to 1970). signaled growth and prosperity for certain groups, and memorialized the continued decimation of others. No matter the period, however, the Census’s evidentiary value concerning the status of people of color has always been significant. When, where, and if people of color are counted depends largely on whether their visibility or invisibility inures to the benefit of white power brokers. 31 See infra section III.B; see also Nathaniel Persily, The Law of the Census: How to Count, What to Count, Whom to Count, and Where to Count Them, 32 Cardozo L. Rev. 755, 764 (2011) (“With respect to reapportionment and redistricting, how one answers the question ‘how to count’ is often highly correlated with one’s answer to the question ‘who stands to benefit.’”). History confirms that people of color are disappeared from the population count unless it is politically or finan­cially expedient to count them and when doing so does not invite political or economic accountability to or for them. 32 As history shows, only Native Americans who were taxed and thus produced reve­nue for the federal government were counted under the Constitution. See U.S. Const. art. I, § 2, cl. 3; id. amend. XIV, § 2. Similarly, in 1787, the South agreed to count a fraction of its population of enslaved Africans mainly to increase the region’s political clout with little attendant financial burden because direct taxes were not routinely levied and enslaved persons were not counted as property. See, e.g., Kenneth N. Addison, “We Hold These Truths to Be Self-Evident . . . ”: An Interdisciplinary Analysis of the Roots of Racism & Slavery in America, at xxii, 281–82, 363 (2009) (“In Article I, Section 2, Clause 3, the Founding Fathers not only legitimized slavery, but also granted a representative bonus in Congress to the slave states by allowing them, for the purpose of representation, to count slaves as three-fifths of a person.”). The modern-day corollary to this exploitation of the confinement of black bodies is prison-based gerrymandering, in which incarcerated persons, who are dis­proportionately black and Latinx, are often counted in rural communities that, in prac­tice, are not politically or financially accountable to them. See, e.g., Julie A. Ebenstein, The Geography of Mass Incarceration: Prison Gerrymandering and the Dilution of Prisoners’ Political Representation, 45 Fordham Urb. L.J. 323, 334–35 (2018) (“Incarcerating people far away from their homes effectively removes residents—disproportionately people of color—from urban areas and credits their voting strength to more rural areas, thus altering the composition of electoral districts and according more power to the votes in districts with prisons and prisoners who cannot vote.”). To the contrary, these communi­ties often benefit financially through local job creation from the warehousing of black and brown bodies in their rural areas. See, e.g., id. at 328–29 (“Rural areas suffering from the loss of farming or manufacturing jobs have seen prisons built in their areas, based, in part, on the promise that prisons provide local economic develop­ment.”); cf. David Gutierrez, Mass Incarceration in Rural Communities: Out of Sight, Out of Mind, Harv. Pol. Rev. (May 27, 2016), [] (“Many of the jobs created by the prison industry only exist during construction and cannot actually be a driving force for sustainable economic growth.”). For more on prison-based gerrymandering, see infra section III.B. The inclusion of the citizenship question is the most recent example of this phenomenon. The  citizenship  question  is  likely  to  reduce  participation  in  the  Census  by  immigrant  and  “immigrant-adjacent” 33 As used in this Essay and in other Census-related works, the term “immigrant-adjacent” refers to the relational proximity to immigrants and immigrant communities by virtue of familial, social, or geographic ties. For example, members of a community who are biologically related to immigrants, maintain social bonds with immigrants, and live in close geographic proximity to immigrants are immigrant-adjacent. See, e.g., 2020 Census: 2019 State Budget & Policy Priorities, N.Y. Immigration Coal., https://
Pagers2019.pdf [] (last visited Feb. 20, 2019) (“The Census Bureau plans to add a citizenship question, which will lower self-response rates from immi­grant and immigrant-adjacent communities.”); see also California v. Ross, 358 F. Supp. 3d 965, 983 (N.D. Cal. 2019) (defining immigrant-adjacent communities as “communities with mixed-status households, where one family member is a U.S. citizen and another family member is not, and communities in which residents would interact with immigrants daily at work, school, or in other similar environments”).
communities—predominantly commu­nities of color—thereby reducing their population count and rendering them invisible to elected representatives. 34 The top five countries of origin for U.S. immigrants are Mexico, China, India, the Philippines, and El Salvador, with Asians forming the largest group of immigrating persons, followed by Latinx persons. See Gustavo López, Kristen Bialik & Jynnah Radford, Key Findings About U.S. Immigrants, Pew Research Ctr.: Fact Tank (Nov. 30, 2018), []. The Census Bureau predicts that in 2030, because the U.S. population is aging, for the first time in history, immigration will produce more population growth than the natural population increase produced by an excess of births over deaths. See Vespa et al., supra note 4, at 10–13. The decreased par­ticipation resulting from inclusion of the citizenship question will also likely deny needed resources to the cities and localities in which those populations concentrate. 35 The Los Angeles County Board of Supervisors wrote to Secretary Ross to explain local conditions in Los Angeles County that would be further complicated by the inclusion of a citizenship question. See Letter from the L.A. Cty. Bd. of Supervisors to Wilbur L. Ross, U.S. Sec’y of Commerce 1 (Jan. 19, 2018) (on file with the Columbia Law Review) (urging the Secretary of Commerce not to include the citizenship question on the short form because it “will likely suppress census response rates by deterring individuals afraid to disclose their immigration status from filling out the census form”). The Board explained to Secretary Ross that the citizenship question would exacerbate the under­count in communities that are already difficult to enumerate:
The State of California, and particularly Los Angeles County, already faces significant challenges in counting minorities, immigrants, and other Hard-to-Survey (HTS) populations. In the 2010 Census, for example, more than 113,000 Latino children in California and an esti­mated 47,000 Latino children in Los Angeles County were not counted, according to a study conducted by the National Association of Latino Elected and Appointed Officials Educational Fund and Child Trends’ Hispanic Institute. Adding a question regarding citizenship on the 2020 Census will make counting Latinos, other migrant communities, and all other HTS populations in the County an even more difficult task.

It is unsurprising that enumeration in the United States is so fraught: The origins of the tensions underlying how, whether, and where to count certain groups in our society are rooted in a history of slavery and anni­hilation. Indeed, these tensions were memorialized in the text of the first Article of the Constitution. Article I, Section 2, Clause 3, which estab­lished the decennial census and the terms of apportioning congressional representatives, presidential electors, and direct taxes, also decreed the erasure of broad swaths of the population on the basis of race:

Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons. The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct. 36 U.S. Const. art. I, § 2, cl. 3 (emphasis added).

Thus, one of the first pronouncements of the U.S. Constitution expressly excluded from enumeration Native Americans who were not subject to taxation, as well as 40% of the population of enslaved Africans by virtue of the Three-Fifths Clause, which counted only three-fifths of every black human in slave-holding states. This constitutionally man­dated “undercount” endured for over eighty years for African Americans 37 It was not until 1868 with the passage of the Fourteenth Amendment, which repealed the Three-Fifths Clause, that African Americans were counted as whole persons. See U.S. Const. amend. XIV, § 1 (“All persons born or naturalized in the United States, and subject to the jurisdiction thereof, are citizens of the United States and of the state wherein they reside.”). and into the first quarter of the twentieth century for Native Americans. 38 Importantly, no Native persons, taxed or untaxed, were considered citizens until the passage of the Indian Citizenship Act of 1924. 8 U.S.C. § 1401(b) (2012). The Act conferred U.S. citizenship on all noncitizen “Indians” born within the territorial limits of the United States. Id. As a result of what came to be known as the “three-fifths compromise”—an accommodation to appease Southern states, which occupied approxi­mately 38% of the seats in the Continental Congress of 1783—the political representation of the South increased to nearly 45% of the seats in the inaugural U.S. Congress seated in 1790. 39 See The Three-Fifth Compromise, Dig. History,
disp_textbook.cfm?psid=163&smtID=3 [] (last visited Feb. 17, 2019).

Counting enslaved Africans, however, did not make Southern gov­ernments accountable to these excluded populations. 40 Despite being counted for purposes of representation in the House of Representatives and the Electoral College, African Americans remained formally enslaved and deprived of basic human rights until the passage of the Thirteenth Amendment in 1865, U.S. Const. amend. XIII, and formally deprived of the franchise until the passage of the Fifteenth Amendment in 1870, under which male inhabitants were granted the right to vote, id. amend. XV. In short, the South benefitted from the partial enumeration of people that it continued to treat as property while avoiding taxation on the full measure of such “property” by counting only a fraction of their being. Rather, counting enslaved Africans—even partially—cemented their continued bondage by creating outsized political dominance in allowing slaveholding states to reduce their poten­tial tax liability 41 See id. art. I, § 2, cl. 3. based on the number of actual residents while augment­ing their political power for representative-apportionment purposes. 42 See id. amend. XIV, § 2. Indeed, Article I of the U.S. Constitution established direct taxation of the States based on total population rather than property. As a result, Southern states had the advantage under the Three-Fifths Clause of enhancing their population for purposes of repre­sentation by counting enslaved Africans, even if only partially. By the same token, Southern states were not subject to taxation on the entire population of enslaved persons and were not politically accountable to them. 43 See id. art. 1, §2, cl. 3.

Moreover, when the Three-Fifths Clause was repealed in 1868 by ratification of the Fourteenth Amendment, it was replaced by a clause that maintained the denigration and discounting of Native Americans by continuing to ex­clude them from the population count for apportion­ment: “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.” 44 Id. (emphasis added). The 1870 U.S. Census showed that 92% of Native Americans were “untaxed” and thus excluded from the popu­la­tion count for purposes of congressional apportionment. 45 See James P. Collins, Native Americans in the Census, 1860–1890, Prologue Mag. (Summer 2006), [] (indicating that, in 1870, only 25,731 of 313,712 total Native Americans were taxed and thus enumerated). As noted below, that same Census also severely undercounted newly emancipated black people. 46     See infra note 132 and accompanying text. Thus, from its inception, the Census has rein­forced to varying degrees the erasure and exclusion of people of color as well as an inequitable distribution of political power and resources premised on their undercount.

II. The Current Stakes

In addition to disappearing certain com­mu­nities of color and other marginalized groups, the Census’s historical undercount of people of color results in compounded intergenerational privations. 47 At least one scholar-journalist has argued that, “[o]f all the ways democracy is threat­ened under President Donald Trump . . . an unfair and inaccurate census could have the most dramatic long-term impact.” Ari Berman, Hidden Figures: How Donald Trump Is Rigging the Census, Mother Jones (May/June 2018), []. When a community is undercounted, fewer resources are allocated to serve its actual population. Its political representation is also diminished because its representatives serve larger constituencies than do representatives of accurately enumerated districts. Over time, the cumu­lative reduction in resources and representation entrenches societal in­equity. These facts compel a course reversal in Census administration to ensure greater fairness toward communities that have been perpetually undercounted rather than the adoption of new policies that threaten to compromise the Census’s accuracy. In addition, the tangible and critical uses to which Census data are put—rights enforcement, allocation of financial and political resources, demography—raise the stakes of continued numeri­cal inaccuracy in counting communities of color.

A. Rights Enforcement

Outside of the reapportionment cases of the 1960s, 48 See, e.g., Reynolds v. Sims, 377 U.S. 533 (1964); Wesberry v. Sanders, 376 U.S. 1 (1964); Baker v. Carr, 369 U.S. 186 (1962). perhaps no legal challenge better underscores the powerful role of the Census in allo­cating political power and enforcing constitutional rights than a little-known suit resolved just before the passage of the Voting Rights Act of 1965, Lampkin v. Connor. 49 239 F. Supp. 757 (D.D.C. 1965). The named plaintiff, Daisy Elizabeth Adams Lampkin, was a renowned African American civil rights activist and suf­fragist who, along with fourteen other eligible voters, sued the Secretary of Commerce and the Census Bureau Director to enforce Section 2 of the Fourteenth Amendment. 50 The Lampkin plaintiffs consisted of two groups of voters. The first group com­prised eligible voters from states that the plaintiffs asserted would gain a congressional seat if the Census appropriately excluded from the apportionment count persons in other states in numbers corresponding to those whose right to vote was denied or abridged in violation of Section 2 of the Fourteenth Amendment. Id. at 757–58. The second group consisted of eligible voters who were denied the right to vote based on a poll tax, literacy test, signature requirement, or other device. Id. at 758. The suit sought to leverage the Fourteenth Amendment’s Penalty Clause to require the Census, and the Secretary of Commerce who administers the Census, to collect data on voting discrimination which would ultimately The plaintiffs insisted that the defendants “compile figures as to denial and abridgment of the right to vote and to prepare and transmit a statement showing the number of Representatives to which each State is entitled on the basis of such figures.”[footnote]Id. Section 2 provides that states that abridge the right to vote shall have their repre­sentation in the House of Representatives reduced in proportion to the number of disenfranchised persons. 51 U.S. Const. amend. XIV, § 2. This Section provides that:
[W]hen the right to vote at any election for the choice of electors for President and Vice-President of the United States, Representatives in Congress, the Executive and Judicial officers of a State, or the members of the Legislature thereof, is denied to any of the male inhabitants of such State, being twenty-one years of age, and citizens of the United States, or in any way abridged, except for participation in rebellion or other crime, the basis of representation therein shall be reduced in the proportion which the number of such male citizens shall bear to the whole number of male citizens twenty-one years of age in such State.
Specifically, the suit demanded a declaration that the Permanent Apportionment Act of 1929—which clarified the methodology of the Census and reappor­tionment and estab­lished the 435-member House of Representatives that exists today 52 2 U.S.C. § 2a (2012). —either imposed a duty to collect sufficient data to enforce the Constitution or was itself unconstitutional. 53 Lampkin, 239 F. Supp. at 759. While dismissed on stand­ing grounds, 54 Id. at 766. this case was among a small handful of lawsuits that sought to enforce the Fourteenth Amendment’s Apportionment Clause as a means of securing equal voting rights. 55 See id. (citing cases in which federal courts of appeal failed to recognize an en­force­able mandate under Section 2 of the Fourteenth Amendment to remedy vote abridge­ment or denial); see also Gerard N. Maggliocca, Our Unconstitutional Reapportionment Process, 86 Geo. Wash. L. Rev. 774, 800–01 & n.147 (2018) (“There were a few cases brought by individuals that unsuccessfully sought to raise a Section Two claim in a tangential way.”). These cases, which were ultimately supplanted by causes of action under the Voting Rights Act of 1965 and its amend­ments, underscore the power of Census data both to carry out the most basic and foundational constitutional de­mands and to provide evidence of discrimination.

As Lampkin demonstrates, because an accurate population count is integral to the equitable distribution of resources and the enforcement of constitutional and civil rights, it is critical that we get it right. Failure to reapportion accurately, for example, is discernable only through an accurate Census count. Indeed, Census data are the foundation of many vot­ing rights claims as well as reapportionment and redistricting litigation. 56 See, e.g., Persily, supra note 31, at 765–70 (discussing the methodology of using Census data in reapportionment and redistricting litigation). Despite the passage of critical civil rights legislation such as the Voting Rights Act, the Fair Housing Act of 1968, 57 42 U.S.C. §§ 3601–3631 (2012). and Title VII of the Civil Rights Act of 1964, 58 Id. §§ 2000e–2000e-17. civil rights enforcement would be nearly impos­sible without accurate Census data. 59 This is not to suggest that current Census data are accurate. Indeed, because of both under- and overcount in the administration of the Census, civil rights litigators often resort to alternative population-count methods in enforcing critical voting rights protec­tions under the Voting Rights Act. See, e.g., Christopher S. Elmendorf & Douglas M. Spencer, Administering Section 2 of the Voting Rights Act After Shelby County, 115 Colum. L. Rev. 2143, 2166, 2187, 2193 (2015) (describing the use of Census sampling to determine threshold factors in Section 2 litigation such as “minority political cohesion” and “racially polarized voting” and to assess “the correlation between race and vote choice”). Such data enable enforcement of the Fair Housing Act’s imperative to affirmatively further fair housing and end segregation 60 See, e.g., Leah Hendey & Mychal Cohen, Urban Inst., Using Data to Assess Fair Housing and Improve Access to Opportunity: A Guidebook for Community Organizations 1 (2017), [] (explaining how community groups and local governments can use Census data to “engage in fair housing and other public planning processes”). and are critical to claims of systemic discrimination in many other statutes. 61 See, e.g., Equal Credit Opportunity Act, 15 U.S.C. § 1691 (2012); Equal Educational Opportunities Act, 20 U.S.C. §§ 1701–1758 (2012); Civil Rights of Institutionalized Persons Act, 42 U.S.C. § 1997. A variety of state and local anti-discrimina­tion statutes also rely on Census data for enforcement. Census data are integral to formulating remedies in education, employment, and public accommodations cases, as well as to advancing claims of intentional discrimination and disparate impact. 62 See Brief of Amicus Curiae NAACP Legal Defense & Educational Fund, Inc., supra note 15, at *25–29 (describing various uses of Census data to promote racial justice and protect against racial discrimination in the areas of political representation, employ­ment, education, housing, and criminal justice); see also Hendey & Cohen, supra note 61, at 20 (stating that “[t]he most widely used sources of demographic data are the Decennial Census and the American Community Survey (ACS)” in describing the use of data in fair housing and access to opportunity analysis); Why We Ask, U.S. Census Bureau, [] (last visited Feb. 17, 2019) (explaining that Census data are used for anti-discrimination mandates, including Civil Rights Act and Voting Rights Act enforcement). And, as Lampkin further illus­trates, there may be ways to leverage the Census to fulfill other constitu­tional mandates, including the protection of equal voting rights.

B. Allocation of Federal Funds

Census data also directly inform the equitable allocation of federal funds that serve as the lifeblood of countless federal, state, and local programs. According to the Census Bureau, “132 programs used Census Bureau data to distribute more than $675 billion in funds during fiscal year 2015.” 63 Marisa Hotchkiss & Jessica Phelan, U.S. Census Bureau, Uses of Census Bureau Data in Federal Funds Distribution: A New Design for the 21st Century 8 (2017),
/Uses-of-Census-Bureau-Data-in-Federal-Funds-Distribution.pdf [].
Those programs included Head Start, Supplemental Nutrition Assistance Program, Section 8 Housing Choice Voucher, Highway Planning and Construction, Federal Pell Grant Program, and Wildlife Restoration, among others. 64 See id. at 3–7 tbl.1. In addition, many programs that provide
vital resources to emerging majority 65 The term “emerging majority” refers to the combined U.S. population of African Americans, Latinx, Asian Americans and Pacific Islanders, and Native Americans that will soon surpass the current total population of the white majority. See Joy Cushman et al., New Org. Inst. Educ. Fund, Engaging the Emerging Majority: The Case for Voter Registration in 2012 and Beyond 2 (2011), [] (defin­ing “[e]merging [m]ajority” as “African Americans, Latinos, Asian/Pacific Islanders, Native Americans and unmarried women” who now “form the majority of the American citizen voting age population”); see also John B. Judis & Ruy Teixeira, The Emerging Democratic Majority 37–39 (2002) (defining in political terms the emerging majority of professionals, working women, and minorities). communi­ties of color are federally funded. 66 See supra note 7.

The failure to accurately enumerate, therefore, has appreciable consequences. When the Census Bureau failed to enumerate 1.5 million California residents in 2010, the California Department of Finance esti­mated that the state lost more than $1.7 billion per year due to the under­count. 67 Gary D. Bass & Adrien Schless-Meier, An Insidious Way to Underrepresent Minorities, Am. Prospect (Nov. 5, 2015), []. The undercounted—who are disproportionately racial, ethnic, and language minorities; children; poor persons; and persons with disabilities—typically require the most services and resources funded by federal dollars. 68 See 2020 Census Fact Sheet, Census Accuracy and the Undercount: Why It Matters; How It’s Measured, Funders Census Initiative 2020, [] (last visited Feb. 17, 2019) (“Demographic analysis and Post-Enumeration Surveys both show that the census misses racial and ethnic minorities, low-income households (indicated by the proxy of homeowner vs. renter), and children (especially ages 0-4), at disproportionately high rates.”). Accordingly, the misallocation of federal funds based on inaccurate or skewed data disproportionately harms the most vulnerable and marginalized communities, leaving state and local governments unable to adequately respond to their constitu­ents’ needs.

C. Reapportionment and Redistricting

As indicated above, perhaps the most critical uses to which Census data are put are apportionment and redistricting. The shifts in congres­sional representation that follow each Census and the increasingly con­tentious redistricting processes to redraw district lines based on new population and reapportionment data demonstrate that Census data are valuable political currency. For example, each of the past two decennial cen­suses resulted in a shift of twelve seats in the U.S. House of Representatives, affecting eighteen states in both 2000 and 2010. 69 Kristin D. Burnett, U.S. Census Bureau, Congressional Apportionment: 2010 Census Briefs 4 (2011), []. In  2010,  eight  states  gained  seats,  including  Texas,  which  gained  four,  and  Florida,  which  gained  two. 70 Id. In both Texas and Florida, a surge in their respective Latinx populations was a substantial cause of the gain. In Texas, people of color accounted for 89% of the state’s growth from 2000 to 2010, with Latinx persons comprising 65% of the state’s growth during this period. 71 Ross Ramsey et al., Minorities Drove Strong Texas Growth, Census Figures Show, Tex. Trib. (Feb. 18, 2011), []. In Florida, Latinx communities similarly contributed to the state’s popula­tion growth relative to others, enabling the state to secure addi­tional congressional seats. 72 See Redistricting in Florida After the 2010 Census, Ballotpedia,
Redistricting_in_Florida_after_the_2010_census [] (last visited Jan. 30, 2019) (“According to an analysis of census data by the Pew Hispanic Center, a major reason why Florida gained two new seats was due to the increase in its Hispanic population.”).
To a lesser degree, similar trends under­pinned the one-seat gains in several of the six other states whose delega­tions increased, including Arizona, Georgia, Nevada, and South Carolina. 73 See Am.’s Voice Educ. Fund, The New Constituents: How Latino Population Growth Will Shape Congressional Apportionment After the 2010 Census 5 (2009), [] (“Without Latinos, Arizona, Florida, Nevada, Oregon, South Carolina and Texas would each have failed to gain as many seats as they are projected to gain.”). This is unsur­prising considering that Latinx persons made up 56% of the total population growth nationwide in the last decade. 74 Jeffrey S. Passel et al., Pew Hispanic Ctr., Pew Research Ctr., Census 2010: 50 Million Latinos; Hispanics Account for More than Half of Nation’s Growth in Past Decade 1 (2011), [].

Notably, these shifts have substantially increased Southern states’ congressional repre­sentation as a result of migration trends from the Northeast toward the South and Midwest. While the South enjoyed just over 30% of House seats for six consecutive decades from 1910 to 1970, by 2010 its share had increased to 37%; after the 2010 Census, “the South . . . maintain[ed] the largest share of House seats among all four regions, as it has since 1940,” 75 Burnett, supra note 70, at 5. and, given the shifts noted above, the South is likely to maintain this share after the 2020 Census. Despite fueling the nation’s growth from 2000 to 2010, however, Latinx commu­nities have not seen commensurate political representation in terms of the number of congressional districts they control or influence. 76 See Am.’s Voice Educ. Fund, supra note 74, at 10 (observing that “[t]he states likely to gain political power following the 2010 Census are currently largely Republican-dominated at both the Congressional and state levels”). “[I]n many cases [states] will owe this expanded power to Latinos who moved to their states over the last several years. Ironi­cally, many members of the delegations who will benefit from the increase in Latino population have embraced policies that are hostile to Latinos and immigrants.” Id. While due in part to voting eligibility criteria that exclude most noncitizens, this is also a function of political margin­al­ization. 77 See Nate Cohn, Why Hispanics Don’t Have a Larger Political Voice, N.Y. Times: Upshot (June 15, 2014), (on file with the Columbia Law Review) (noting that “half of all Hispanics live in just 65 of the nation’s 435 congressional districts”). For example, Texas’s redistricting cycle following the 2010 Census resulted in multiple lawsuits finding that Texas intentionally discriminated against its Latinx residents in the redistricting process. 78 In 2011, civil rights groups commenced a lengthy challenge to Texas’s post-2010 re­districting plan. See Jim Malewitz & Alexa Ura, How the Texas Redistricting Lawsuit Outlived a Voter Who Sued, Tex. Trib. (Oct. 18, 2017),
drawn-out-texas-redistricting-fight-people-are-passing-away-while-awai/ []. Despite Latinx persons constituting over 38% of the state’s population and increasing the state’s population count relative to other states and other groups, see State and County QuickFacts: Texas, U.S. Census Bureau, [] (last visited Feb. 17, 2019), the maps diluted and undervalued Latinx political representation. At the end of the last Term, however, the Supreme Court held that ten of the eleven districts that Latinx people challenged as racially discriminatory were constitutional. See Abbott v. Perez, 138 S. Ct. 2305, 2335 (2018). The only district that the Court struck down as a racial gerrymander was House District 90, which includes the city of Fort Worth, because there was incontrovertible evidence that the legislature unlawfully used race as the predominant factor in drawing the district’s boundaries. Id. at 2334–35.
In Abbott v. Perez, after reversing several other claims, the Supreme Court affirmed the district court’s ruling that Texas violated the Fourteenth Amendment by using race as “the pre­dominant factor in the design” of House District 90, which relied on the manipula­tion of both African American and Latinx communities, resulting in “an impermissible racial gerrymander.” 79 138 S. Ct. at 2334–35.

By all indications, minority population growth—and especially Latinx population growth—will result in similar apportionment consequences following the 2020 Census. 80 Congressional apportionment also impacts the composition of the Electoral College and, therefore, who is elected President. See Distribution of Electoral Votes, U.S. Electoral College, Nat’l Archives & Records Admin., [] (last updated Dec. 10, 2010). Electoral College votes are allocated based on total state population and congressional delegations. U.S. Const. art. II, § 1, cl. 2 (“Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in Congress . . . .”). Because of the timing of the 2020 Census, the composition of the Electoral College that will elect the President in 2020 will be unaffected by new Census re­sults. See Distribution of Electoral Votes, supra (noting that the current allocations based on the 2010 Census are effective for the 2012, 2016, and 2020 presidential elections). Election Data Services has released an apportionment study based on 2018 Census data that predicts an increase or decrease in the congressional delegations of thirteen states as a result of the 2020 Census; 81 See Press Release, Election Data Servs., Arizona Gains Rhode Island’s Seat with New 2018 Census Estimates; But Greater Change Likely by 2020, at 1 (Dec. 19, 2018) [hereinafter Election Data Servs. Press Release],
wp-content/uploads/2018/12/NR_Appor18.pdf [].
this change is fueled again largely by Latinx population growth. 82 See, e.g., Jesse Alston, The Rust Belt Elevated Trump, but Its Electoral Power Is Dwindling, FiveThirtyEight (Feb. 16, 2017), [] (referencing “increasing Latino political power” evidenced by the prediction that “[s]even of the nine seats gained will go to states where more than 1 in 5 residents is Latino, and all nine will go to states where at least 15 percent of minors are Latino”). Further, “Latinos made up more than half of the United States’ population growth from 2000 to 2014, and they’re especially concentrated in states that will likely gain electoral votes in 2024.” Id. Specifically, the states of Arizona, Colorado, Florida, North Carolina, and Oregon may each gain a single seat and Texas will likely net two additional seats. 83 Election Data Servs. Press Release, supra note 82, at 1. Each of these states has enjoyed a sizeable increase in its Latinx population since 2000. 84 See Renee Stepler & Mark Hugo Lopez, Ranking the Latino Population in the States, Pew Research Ctr.: Hispanic Trends (Sept. 8, 2016), [] (showing that, by 2014, Arizona, Colorado, Florida, and Texas had Latino populations of at least one million and that North Carolina’s Latinx population increased 136% be­tween 2000 and 2014); see also Casey Parks, Oregon’s Latino Population Is Growing Faster than the Nationwide Average, Oregonian: OregonLive (Sept. 15, 2016), [] (reporting that Oregon’s Latinx population “is growing faster than the national rate and is significantly younger than others in the state” and that the number of Latinx “increased by 72 percent since 2000, compared to only 50 percent nationwide”).

It is difficult to quantify the precise impact on apportionment that the anticipated undercount—resulting from inclusion of a citizenship question and other threatened barriers to participation—would have. It  is  beyond  doubt,  however,  that  an  undercount  will  undermine  the  one-person,  one-vote  principle  and  endanger  the  counting  of  the  total  popu­lation  for  purposes  of  both  apportionment  and  redistricting. 85 See, e.g., New York v. U.S. Dep’t of Commerce, 351 F. Supp. 3d 502, 594 (S.D.N.Y. 2019) (finding, based on expert testimony, that as a result of the inclusion of the citizen­ship question, “several states, including some that are Plaintiffs here and some in which NGO Plaintiffs’ members reside, will lose at least one seat in the congressional reappor­tionment based on the 2020 census data”), cert. granted, 139 S. Ct. 953 (2019) (mem.). Nothing under­scores this more than the battle before the Supreme Court in 2016 challenging the use of total population count in redistricting. In Evenwel v. Abbott, residents of rural Texas counties argued that their votes were not equal to votes from districts with more nonvoters. 86 See 136 S. Ct. 1120, 1123 (2016). The plaintiffs claimed that voters in districts with fewer nonvoters received less repre­sentation than voters in districts with a larger number of nonvoters, thus violating the Fourteenth Amendment. 87 See id. at 1123, 1125. Under the plaintiffs’ theory, elected officials are primarily accountable to their eligible-voter constit­uents, rather than to nonvoters, in their districts. As a result, according to the plaintiffs, counting total population for apportionment purposes results in districts with higher numbers of nonvoters receiving more representa­tion than voters in districts with fewer nonvoters. 88 See id. at 1125–26. Accord­ingly, plaintiffs sought to ban the use of total population in redistricting and to count only eligible voters for purposes of determining the congres­sional repre­sentation of each state.

In a unanimous decision authored by Justice Ginsburg, the Court upheld the practice of counting all residents for purposes of representa­tion under Article I, Section 2 of the Constitution. 89 Id. at 1123. The Court relied on records from constitutional conventions in which the Framers advocated vigorously for an inclusive and expansive view of appor­tionment, as well as Justice Ginsburg’s definition of the constitutional theory of “represen­tational equality,” which deems all residents of the state to be constitu­ents deserving of enumeration and representation. 90 See id. at 1126–27, 1129–32; see also Nelson, supra note 22 (“Justice Ginsburg masterfully unpacks the constitutional history of the principle of counting all residents for purposes of representation grounded in Article I, Section 2 of the Constitution, as well as in the Constitutional Conventions in which framers . . . powerfully advocated for an inclu­sive and expansive view of apportionment.”). In this context, representational equality is a theory of democracy that accounts for and recognizes the entitlement of every human within the nation’s borders to have representation. 91 See Evenwel, 136 S. Ct. at 1129 (“[I]t remains beyond doubt that the principle of representational equality figured prominently in the decision to count people, whether or not they qualify as voters.”). The concept of “equal representation” or “representational equality” derives from the one-person, one-vote cases, which refer to the principle of “equal representation for equal numbers of people.” See, e.g., Kirkpatrick v. Preisler, 394 U.S. 526, 530–33 (1969) (citing “the constitutional command to provide equal representation for equal numbers of people”); Wesberry v. Sanders, 376 U.S. 1, 18 (1964) (describing the “Constitution’s plain objective of making equal representation for equal numbers of people the fundamental goal” in drawing congressional districts). Former Ninth Circuit Court of Appeals Judge Alex Kozinski observed in his dissent in Garza v. County of Los Angeles that the one-person, one-vote principle can be interpreted in two different ways: (1) as a “principle of equal representation,” whereby the total number of residents in each district must be equal such that “constituents have more or less equal access to their elected officials, by assuring that no official has a disproportionately large number of constituents to satisfy”; or (2) as a “principle of electoral equality” which “recognizes that electors—persons eligible to vote—are the ones who hold the ultimate political power in our democracy.” 918 F.2d 763, 781 (9th Cir. 1990) (Kozinski, J., concurring in part, dissent­ing in part). This view is also consonant with Section 1 of the Fourteenth Amendment, which cured the ignoble three-fifths compro­mise, counting African Americans as citizens before African American males were given the right to vote in 1870 92 U.S. Const. amend. XV. —and before any women were given the right to vote in 1920. 93 Id. amend. XIX.

To require that each state’s allocation of districts for the U.S. House of Representatives be based on total population but then permit the bound­aries of those districts to be drawn using only voter population would create an “apportionment anomaly” and upend apportionment and dis­tricting practices with longstanding historical roots. 94 There was overwhelming bipartisan support among state and local governments for continued reliance on total population for redistricting purposes. Twenty-one states signed an amicus brief in support of total population count, including ten states with Republican governors and eleven states with Democratic governors. See Brief for the States of New York, Alaska, California, Delaware, Hawai’i, Illinois, Indiana, Iowa, Maine, Maryland, Massachusetts, Michigan, Minnesota, Mississippi, New Hampshire, North Dakota, Oregon, Rhode Island, Vermont, Virginia, & Washington as Amici Curiae in Support of Appellees at 1, Evenwel, 136 S. Ct. 1120 (No. 14-940), 2015 WL 5719756. For example, the Court has reinforced the principle of equipopulous districts based on total population count for over fifty years. The “reappor­tionment revolution,” as some have aptly named it, began with an extreme racial gerrymander in Alabama that was successfully challenged in Gomillion v. Lightfoot. 95 See 364 U.S. 339, 347 (1960). Subsequently, in Baker v. Carr, the Supreme Court held that a complaint challenging Tennessee’s gross malapportionment presented a justiciable constitutional cause of action, irreversibly enter­ing the Court into the political thicket of redistricting cases. 96 See 369 U.S. 186, 237 (1962). These  landmark  deci­sions  laid  the  foundation  for  the  seminal  case  Reynolds v. Sims,  which  estab­lished  the  principle  of  one-person,  one-vote  and  the  ensuing  require­ment  that  con­gressional districts  be  populated  as  equally  as  possible. 97 See 377 U.S. 533, 568 (1964).

By rejecting the proposition in Evenwel that only certain segments of the population should be counted for redistricting purposes, the Court protected vulnerable populations that would have been most harmed by the exclusion of persons ineligible to vote, including the seventy-five mil­lion children in the United States (thirteen million of whom are African American), immigrants on the path to citizenship, citi­zens denied the right to vote due to a criminal conviction, and persons with mental disabilities. 98 Researchers estimated that over twenty million African Americans were at risk of exclusion from representation if total population Census data were rejected as the appro­priate population data for redistricting. See Background on Evenwel v. Abbott, ACLU,
sheet.pdf [] (last visited Feb. 17, 2019).
Thus, it is critically important—for both apportionment and redistricting—that the Census, whose data are fundamental to apportionment, accurately enumerate the population and reflect its unprece­dented changing demographics. Furthermore, accurate Census data enable representational equality, in which all per­sons are “counted—and served—as constituents,” 99 Nelson, supra note 22. to take hold.

D. Changing Demographics

Almost ten years ago, the 2010 Census revealed a U.S. population that was nearly 40% people of color. Specifically, Latinx persons made up 16.3% of the total population, Asians made up 4.8%, African Americans made up 12.6%, Native Americans made up 0.9%, and an additional 2.9% of the population identified as two or more races. 100 Karen R. Humes et al., U.S. Census Bureau, Overview of Race and Hispanic Origin: 2010, at 4 tbl.1 (2011), []. The projections for 2030 reveal a further browning of America. The number of Latinx persons is expected to eclipse that of all other racial and ethnic groups by increas­ing nearly 5% to occupy 21.1% of the population. 101 See Population Div., U.S. Census Bureau, Projected Race and Hispanic Origin: Main Projection Series for the United States, 2017–2060 (2018), (on file with the Columbia Law Review) (providing projected population statistics based on race and Hispanic origin); see also Vespa et al., supra note 4, at 7 tbl.3 (projecting population by race and ethnicity at 2030 and at 2060). The Census Bureau did not find “a statistically significant undercount for the Asian or for the Native Hawaiian and Other Pacific Islander populations in 2010 (at 0.1 and 1.3 percent, respectively). These estimates were also not statistically different from the results measured in 2000 (a 0.8 percent overcount and a 2.1 percent undercount, respectively).” Census Bureau Releases Estimates of Undercount and Overcount in the 2010 Census, U.S. Census Bureau (May 22, 2012), [] [hereinafter Census Estimates of Undercount and Overcount]. In addition, the Census Bureau noted the effect of geography on the counting of the American Indian and Alaska Native population, finding that “American Indians and Alaska Natives living on reservations were undercounted by 4.9 percent, compared with a 0.9 percent overcount in 2000.” Id. Asians will also significantly increase in number with a projected population share of 6.9%. 102 Vespa et al., supra note 4, at 7 tbl.3. Census projections predict that in 2030, African Americans will comprise 13.8% of the population while Native Americans and those who identify as two or more races will be 1.3% and 3.6%, respectively. 103 Id. Coupled with demographics that show that the U.S. child population is already nearly majority-minority, the multicultural trajectory of America is ineluctable. 104 See Child Population by Race in the United States, The Annie E. Casey Found. KIDS COUNT Data Ctr.,,870,573,869,36,868,867,133,38,35/68,69,67,12,70,66,71,72/423,424 [] (last visited Jan. 30, 2019) (presenting data show­ing that, in 2017, non-Hispanic white children made up 51% of the child population in the United States).

If administered correctly, the 2020 Census will be the first to etch a portrait of a new America whose future population is irreversibly racially and ethnically diverse and precipitously less white. And it should be lost on no one that the manifold government-created challenges to administer­ing a fair and accurate Census—underfunding, the failure to appoint a director who will have ample time to prepare and perform needed course correction, and the cancellation of vital tests on new technology and methodologies—coincide with this historic demographic turning point. 105 John Thompson, who spent more than thirty years at the Census Bureau and was its director from 2013 to 2017, submitted court testimony that Secretary Ross’s decision to include a citizenship question was “unprecedented.” Rule 26(A)(2)(B) Expert Report & Declaration of John Thompson at 8, N.Y. Immigration Coal. v. U.S. Dep’t of Commerce, No. 1:18-cv-02921-JMF (S.D.N.Y. Jan. 15, 2019), 2018 WL 6976798. Thompson further stated that “he never observed a political official at the Department of Commerce solicit another federal agency to request that a specific question be added to the Decennial Census questionnaire.” Id. Underfunding the Census will result in an undercount of urban areas in which minority populations tend to concentrate 106 See Kim Parker et al., Pew Research Ctr., What Unites and Divides Urban, Suburban and Rural Communities 4, 16–17 (2018), [] (noting that “[w]hites have become a minority of the pop­ulation in most urban counties since 2000”). and may also exacerbate the undercount of rural communities. 107 See Arloc Sherman, Underfunding the 2020 Census Would Likely Harm a Diverse Range of Communities, Including Many in Rural America, Ctr. on Budget & Policy Priorities: Off the Charts Blog (Mar. 1, 2018), [] (“Concerns about the accuracy of the upcoming census often focus on people of color, who face a higher risk that they will be undercounted. Less well known is that disadvantaged rural Americans and other low-income and marginalized groups are also at high risk.”). As noted earlier,  “being  undercounted  not  only  would  deprive  [a community]  of  fair  representation  in  Congress  and state  capitols  and  their  fair  share  of  federal  dollars, it  also  would  distort the information that a business uses to decide where to locate or a county uses to decide whether to close a school.” 108 Id. Therefore, the moral and historical imperative to capture this moment of demographic revolution in the most influential modern de­moc­racy with precision and care counsels against changes to the administration of the Census that would exacerbate an undercount, espe­cially the inclusion of a citizenship question.

The country’s changing demographics, along with the fair and equitable allocation of fiscal and political resources, the ability to enforce civil rights protections with reliable data, and the fraught history of (dis)counting people of color that lies at the root of our democracy’s Founding, underscore the urgent need for accuracy in enumerating an increasingly diverse society. 109 For example, this issue is currently playing out along the spectrum of gender identity. In a preliminary report issued by the U.S. Census Bureau’s Center for Survey Measurement Research and Methodology Directorate, self-identified transgender survey respondents expressed widespread support for including gender-identity questions on the Census but also expressed significant concern about the purposes for which the data would be used, confidentiality, and the feasibility of data collection. See Jessica L. Holzberg et al., Ctr. for Survey Measurement, U.S. Census Bureau, Assessing the Feasibility of Asking About Gender Identity in the Current Population Survey: Results from Focus Groups with Members of the Transgender Population 21–28 (2018), [https://]; see also Samantha Allen, Gay and Single? Bisexual? Transgender? The 2020 Census Still Erases You, Daily Beast (Apr. 3, 2018), [
XX6Z] (“Although gaining a more accurate estimate of the number of same-sex couples will be an important milestone for both research and policymaking, the 2020 Census will still leave out most bisexual people, unpartnered gay men and lesbians, and transgender people— . . . the vast majority of the LGBT community.”).
It is critical, therefore, to identify the most significant barriers to an exact count and potential mitigating prescriptions.

III. Barriers to an Exact Count

Many challenges beset the 2020 Census. After abruptly quitting his post just four months following the change in presidential administra­tion, former Census director John Thompson confessed that “[t]here’s a lot of concern about the potential for an undercount in the [C]ensus.” 110 Alvin Chang, How Republicans Are Undermining the 2020 Census, Explained with a Cartoon, Vox (Aug. 30, 2018), []. The $15.6 billion price tag to administer the 2020 Census 111 U.S. Gov’t Accountability Office, GAO-18-543T, 2020 Census: Actions Needed to Mitigate Key Risks Jeopardizing a Cost-Effective and Secure Enumeration 3 (2018) [hereinafter GAO 2020 Census Report], []. —$3.3 billion higher than the $12.3 billion cost of the previous Census—combined with changes in both the technology and method­ology of the Census have left the process vulnerable to a high risk of failure. 112 See Teresa Mathew, The Constitutional Counting Crisis, CityLab (Oct. 13, 2017), [] (“Come 2020, the U.S. Census Bureau will face a perfect storm of logistical nightmares: reduced funding, poor planning, and a climate of fear that may exacerbate the challenges of tabulating communities that have historically been hard to count.”); see also U.S. Gov’t Accountability Office, GAO 17-317, High Risk Series: Progress on Many High-Risk Areas, While Substantial Efforts Needed on Others 41–45 (2017), [
UNC8-VUXP] (“Over the past 3 years, we have made 30 recommendations to help the Bureau design and implement a more cost-effective census for 2020; however, only 6 of them had been fully implemented as of January 2017.”).
Like their historical predecessors, today’s Census controversies demand that gov­ernment officials make hard choices about accuracy, accountability, and inclusion. The current Administration has made its posi­tion clear by divesting the Census of necessary funds to  properly  administer  the  survey  with  new  technology,  compromising  its infra­struc­ture  and  integrity,  fail­ing  to  appoint  a  new  director  of  the  Bureau,  and  attempting  to  include  a  citizenship  question  while fomenting  an  increasingly anti-immigrant cli­mate. 113 See GAO 2020 Census Report, supra note 112, at 41–45; see also California v. Ross, 358 F. Supp. 3d 965, 975 (N.D. Cal. 2019) (“[A] significant differential undercount, particularly impacting noncitizen and Latino communities, will result from the inclusion of a citizenship question on the 2020 Census, compounded by macro-environmental fac­tors arising out of the national immigration debate.”). This section examines key barriers—including the citizenship question and prison gerrymandering, among others—to an accurate count of the important demographic change in communities of color.

A. The Citizenship Question

There is overwhelming evidence that including a citizenship question on the Census would be detrimental to an accurate count of commu­nities of color. 114 As noted earlier, the Supreme Court will be deciding on the legality of the citizen­ship question shortly after this Essay goes to press. The enormously harmful impact of such a question, however, is worth exploring regardless of the ultimate outcome of the liti­gation as it helps shed light on the numerous barriers that communities of color have faced and continue to face in an effort to be accurately counted. In the first trial to determine the legality of including the citizenship question on the Census, a federal judge found that a preponderance of the evidence, including Secretary Ross’s trial testi­mony, showed “that the citizenship question will cause a significant dif­ferential decline in self-response rates among noncitizen households” as compared to other households. 115 New York v. U.S. Dep’t of Commerce, 351 F. Supp. 3d 502, 583 (S.D.N.Y. 2019), cert. granted, 139 S. Ct. 953 (2019) (mem.). The court also described addi­tional evidence showing that the citizenship question will cause a “net differential decline” in self-response rates among noncitizen households. See id. at 583. This decline in response rates will, in turn, produce a “‘net differential under­count’ of people who live in non­citizen households.” 116 Id. at 578 n.34. A subsequent court found that inclusion of the citizenship question not only violates the Administrative Procedure Act as arbitrary and capricious but also contravenes the Enumeration Clause itself “because its inclusion will materially harm the accuracy of the cen­sus without advancing any legitimate governmental interest.” 117 California, 358 F. Supp. 3d at 1048.

Including a citizenship question would also be a striking aberration from recent Census administration. Since 1940, the government has administered the Census by sending a “short” form to most addresses throughout the country to collect general information about households and their inhabitants. 118 See Hotchkiss & Phelan, supra note 64, at 9. As noted above, the short form has not included a citizenship question in over seventy years. 119 See supra note 8 and accompanying text. The Census long form—which was sent to approximately one in six households in 2000—contains a citizenship question, along with questions concerning educational attainment, disability status, employment sta­tus, income, and housing costs. See U.S. Census Bureau, United States Census 2000 Form D-61B (2000), []; see also Beth Jarosz, Continuity and Change in the U.S. Decennial Census, Population Reference Bureau (Mar. 25, 2018),
in-the-u-s-decennial-census/ [].
Nor  has  the  Census  ever  asked  the  entire  population  it  surveys  whether  each  person  was  a  citizen. 120 Thomas P. Wolf & Brianna Cea, A Critical History of the United States Census and Citizenship Questions, 108 Geo. L.J. Online 1, 13–14 (2019), []. Rather, contrary to the representations of this administration, this radical alteration of the Census would be “historically unprecedented.” 121 Id. at 14.

The Census Bureau’s own studies have also revealed that including a citizenship question may have potentially deleterious effects, including heightening immigrants’ and others’ fear of participating in the Census. 122 Mikelyn Meyers, U.S. Census Bureau, Respondent Confidentiality Concerns and Possible Effects on Response Rates and Data Quality for the 2020 Census 2, 9, 12 (2017), []. For example, Census researchers observed an “increase in respondents . . . spontaneously expressing concerns to researchers and field staff about confidentiality and data access relating to immigration[;] [l]egal resi-dency[;] [and] [t]he perception that certain immigrant groups are unwelcome,” as well as “increased rates of unusual respondent behaviors during pretesting and production surveys (data falsification, item non-response, break-offs).” 123 Id. at 2–3. The report gave as illustrative examples several quotes from respond­ents during pretesting about their fears concerning participation in the Census. One Spanish-speaking respondent noted that “[t]he possibility that the Census could give my information to internal security and immigration could come and arrest me for not having documents terrifies me.” Id. at 8. Another Spanish-speaking respondent empha­sized the current political climate and said that “the Latino community will not sign up because they will think that Census will pass their information on and people can come looking for them.” Id. An English-speaking respondent noted that the “Muslim ban” was a concern. Id. These behavioral changes among Census respond­ents substantiate concerns that a citizenship question would compromise the survey’s accuracy. 124 See id. at 3, 15. Notably, the irregularities in Census responses occurred prior to the announcement of the intent to include a citizenship question on the Census and, presum­ably, these irregularities would only be exacerbated by it. 125 See Henry Goldman, Immigrants Hiding from Trump Imperil Accuracy of U.S. Census, Bloomberg Quint (Mar. 21, 2018),
immigrants-hiding-from-trump-imperil-accuracy-of-u-s-census (on file with the Columbia Law Review) (last updated Mar. 22, 2018) (stating that “[m]illions of foreign-born residents are expected to hide from or avoid the 2020 count because of the political climate created by Trump” and that the inclusion of a citizenship question will have an exacerbating effect); cf. Emily Baumgaertner, Spooked by Trump Proposals, Immigrants Abandon Public Nutrition Services, N.Y. Times (Mar. 6, 2018),
politics/trump-immigrants-public-nutrition-services.html (on file with the Columbia Law Review) (reporting that “[s]tatistics on participation in state and local efforts show fewer people are using an array of food programs . . . and food banks” following the Trump Administration’s threat to consider use of government benefits in evaluating green card or visa applications).

The potential impact of the citizenship question on people of color is evident. Immigrant communities are predominantly composed of people of color: The top five countries of origin for the U.S. immigrant popula­tion are Mexico, China, India, the Philippines, and El Salvador, and today, Asians form the largest group of immigrating persons, fol­lowed by Latinx persons. 126 See López at al., supra note 34. A sizeable number of black U.S. residents are also noncitizens: Roughly 42% of black immigrants do not have U.S. citizen­ship. 127 Monica Anderson & Gustavo López, Key Facts About Black Immigrants in the U.S., Pew Research Ctr.: Fact Tank (Jan. 24, 2018), []. Further, regardless of immigration or citizenship status, black U.S. residents are directly impacted by the threat of a citizenship question due to the significant number of black people in the U.S. who live in or near hard-to-count immigrant and noncitizen communities. 128 See Randy Capps, Kristen McCabe & Michael Fix, Migration Policy Inst., Diverse Streams: Black African Migration to the United States 11 (2012),
sites/default/files/publications/CBI-AfricanMigration.pdf [] (demonstrating that “Black African immigrant popula­tions are heavily concentrated in a handful of states, and their distribution largely reflects the pattern for the total Black population”); John Iceland & Kyle Anne Nelson, The Residential Segregation of Mixed-Nativity Married Couples, 47 Demography 869, 887 (2010) (showing that black immigrant or mixed-immigrant households in metropolitan regions are highly segregated from non-black racial and ethnic groups). In addition, because an estimated 11% of foreign-born Muslims in the United States—and 32% of U.S.-born Muslims—identify as black, Muslims in America: Immigrants and Those Born in U.S. See Life Differently in Many Ways, Pew Research Ctr.: Religion & Pub. Life (Apr. 17, 2018), [https://], as with other Muslim immigrant communities of color, anti-Muslim and anti-immigrant rhetoric compound to amplify fears that could discourage black Census participation.

The anti-immigrant climate in which the unprecedented proposal to include a universal citizenship question arose has heightened the concerns of immigrant and immigrant-adjacent communities. 129 See, e.g., Memorandum from the Ctr. for Survey Measurement on Respondent Confidentiality Concerns to Assoc. Directorate for Research & Methodology 1 (Sept. 20, 2017), [] (“[R]esearchers heard respond­ents express new concerns about . . . the ‘Muslim ban,’ discomfort ‘registering’ other household members by reporting their demographic characteristics, the dissolu­tion of the ‘DACA’ . . . program, [and] repeated references to Immigration and Customs Enforcement (ICE), . . . and reported that respondents’ fears, particularly among immi­grant respondents, have increased markedly this year.”). Harsh rhetoric against immigrants has coincided with increased hostile govern­ment action in the form of immigration enforcement, arrests, and deporta­tion. 130 See, e.g., Kristen Bialik, ICE Arrests Went Up in 2017, with Biggest Increases in Florida, Northern Texas, Oklahoma, Pew Research Ctr.: Fact Tank (Feb. 8, 2018), [] (“After years of decline, the number of arrests made by U.S. Immigration and Customs Enforcement (ICE) climbed to a three-year high in fiscal 2017 . . . .”); Ron Nixon, Immigration Arrests and Deportations Are Rising, I.C.E. Data Show, N.Y. Times (Dec. 14, 2018), (on file with the Columbia Law Review) (noting an 11% increase in arrests and a 13% increase in deportations by Immigration and Customs Enforcement from fiscal year 2017 to fiscal year 2018). Fear of reprisal or even mere concerns about being identified as noncitizens may quell interest in being counted, leading to an increased likelihood of undercount for people of color.

Nor is the threat of undercount manufactured. Apart from a lack of participation, Census administration has historically undercounted people of color who wish to participate. As early as 1870—when the government conducted the first Census after the ratification of the Reconstruction Amendments—black people “were egregiously undercounted . . . but lacked the political clout to secure a recount.” 131 Samuel Issacharoff & Allan J. Lichtman, The Census Undercount and Minority Representation: The Constitutional Obligation of the States to Guarantee Equal Representation, 13 Rev. Litig. 1, 6 (1993). The problem of undercount persists today. Even without a citizenship question, the 2000 Census failed to count approximately one million African Americans. 132 Barry Edmonston, The Annie E. Casey Found. & The Population Reference Bureau, The Undercount in the 2000 Census 3 (2002),
uploads/2002/05/undercount_paper.pdf [].
And, despite the belief that it is one of the most accurate Censuses to date, 133 See Census: Planning Ahead for 2020: Hearing Before the Subcomm. on Fed. Fin. Mgmt., Gov’t Info., Fed. Servs. & Int’l Sec. of the S. Comm. on Homeland Sec. & Governmental Affairs, 112th Cong. 52 (2012) (prepared statement of Robert M. Groves, Director, U.S. Census Bureau) (“[T]his showed that the 2010 census was one of the best (if not the best) censuses this country’s ever seen.”). the 2010 Census overcounted whites by 0.8% and undercounted 2.1% of the black population (the approximate size of two legislative districts) and 1.5% of the Latinx population. 134 Census Estimates of Undercount and Overcount, supra note 102.

The effects of these Census inaccuracies are compounded by the “dif­ferential undercount”: the gap between the overcount of whites and the undercount of communities of color. The differential undercount exac­erbates the harm of undercount because it results in the over-resourcing of overcounted communities. 135 See Funders Comm. for Civic Participation, Census Accuracy and the Undercount: Why It Matters; How It’s Measured 2 (2017), [] (explaining the differential overcount and calculating an approxi­mately 3% differential overcount resulting from the 2010 Census’s 2.06% undercount of the black population and 0.83% overcount of the white population). But see Peter Skerry, We’re Overstating the Importance of the Undercount, Brookings Inst. (May 14, 2000), [] (recognizing that “[t]here is no dis­puting the reality of what the bureau calls ‘the differential racial undercount’” but arguing that its impact is overstated). As outlined below, myriad fac­tors, includ­ing poverty, displacement, and the digital divide, contrib­ute to the under­counting of communities of color, which virtually guar­antees a Census that mischaracterizes the country’s demographics.

The political consequences of undercount are also deeply concern­ing. Some estimates suggest that the inclusion of a citizenship question could produce an undercount of noncitizens as high as 15%. 136 Emily Badger, A Census Question That Could Change How Power Is Divided in America, N.Y. Times: Upshot (July 31, 2018), (on file with the Columbia Law Review). Such an undercount (or anything approaching it­) would increase the congres­sional delegations of Colorado and Montana by one and decrease those of New York and California — which  contain  among  the  highest  concen­trations  of  racial  and  ethnic  minorities  in  the  country ­— by  the  same  amount. 137 See Population Distribution by Race/Ethnicity, State Health Facts, 2017, Henry J. Kaiser Family Found.,–hispanic–asian–american-indianalaska-native–native-hawaiianother-pacific-islander–two-or-more-races–total&sortModel=
%7B%22colId%22:%22Total%22,%22sort%22:%22desc%22%7D [] (last visited Apr. 10, 2019) (listing California, Texas, Florida, and New York as the states with the highest raw numbers of racial minorities, respectively).
Moreover, undercounts of noncitizens benefit Republicans by reducing the distribution of power and resources to the Democratic-leaning areas that noncitizens are more likely to inhabit. 138 See Harry Enten, Blue States Are Far More Likely to Lose Money and Power over Census Citizenship Question, CNN (Mar. 27, 2018), [] (“[T]here are 13 million non-citizens living in states . . . won by Democrat Hillary Clinton in 2016. There are only about 9.5 million non-citizens living in states won by Republican Donald Trump. In other words, about 58% of non-citizens live in Clinton states and 42% live in Trump states.”). Indeed, all inhabitants of immigrant or immigrant-adjacent communities are more likely to be harmed by the undercount of the immigrant and immigrant-adjacent population. 139 See Wendy R. Weiser & Thomas Wolf, Why the Census Asking Citizenship Is Such a Problem, Brennan Ctr. for Justice (Mar. 27, 2018), [] (noting that the impact of including a citizenship question “would be dramatic” and “would espe­cially harm communities with high immigrant populations”).

Finally, aside from the harm that undercounting will cause through its ef­fect on apportionment of political power, the fear that this Administration could use the Census to identify and deport immigrant populations is not unfounded. In  1943,  at  the  height  of  World  War  II,  the  Census Bureau  released  the  names  and  addresses  of  American  residents  of  Japanese  descent  to  facilitate  their  capture  and  internment. 140 See Teresa Watanabe, In 1943, Census Released Japanese Americans’ Data, L.A. Times (Mar. 31, 2007), [] (reporting on the newly released documents confirming that the Census Bureau released the names and addresses of Japanese residents in the United States under wartime legislation, which led to their internment); see also JR Minkel, Confirmed: The U.S. Census Bureau Gave Up Names of Japanese-Americans in WW II, Sci. Am. (Mar. 30, 2007), []. A docu­ment published by the Census Bureau acknowledges that “[t]he histor­ical record is clear that senior Census Bureau staff proactively cooperated with the internment, and that census tabulations were directly implicated in the denial of civil rights to citizens of the United States who happened also to be of Japanese ancestry.” 141 U.S. Census Bureau, A Monograph on Confidentiality and Privacy in the U.S. Census 16 (2001) [hereinafter Monograph on Confidentiality], []. Although the release of this infor­mation was legal under the Second War Powers Act of 1942, it violated the trust of Japanese American citizens who gave information under an expectation that such information would be kept confidential, and it may undermine the public’s trust in the confidentiality of modern Censuses. 142 See Watanabe, supra note 141 (stating that, although “legal under wartime leg­islation,” the disclosures were “arguably unethical” and may have affected “public trust in the bureau’s confidentiality pledges”). In the years following World War II, the Census has maintained certain confidentiality restrictions; 143 See Monograph on Confidentiality, supra note 142, at 17–20 (describing the efforts to ensure the confidentiality of the Census after World War II, including statutory measures and litigation). nonetheless,  government  officials  have  mis­used  Census  data  and  the  potential  of  a  breach  of  said  confi­dentiality  restrictions  is  a  real  risk  in  the  near  future. 144 See Freedom of Information Documents on the Census: Department of Homeland Security Obtained Data on Arab Americans from Census Bureau, Elec. Privacy Info. Ctr., [] (last visited Jan. 30, 2019) (discussing documents obtained by EPIC through a public records request showing that the Census Bureau gave the Department of Homeland Security “statistical data on people who identified themselves on the 2000 census as being of Arab ancestry” and that “special tabulations were prepared specifically for the law enforcement agency”); see also Tierney Sneed, Newly Released DOJ Email Prompts Questions About Census Confidentiality, Talking Points Memo (Nov. 20, 2018), [] (discussing an email exchange that shows the Justice Department obfuscat­ing its approach to determining whether the Patriot Act can compel disclosure of Census data to law enforcement). Confidential government data have also been put to criminal use by government officials. See Press Release, U.S. Dep’t of Justice, Former ICE Chief Counsel Pleads Guilty to Using the Identities of Numerous Aliens for Wire Fraud and Aggravated Identity Theft Scheme (Feb. 15, 2018), [] (“Former Chief Counsel Raphael A. Sanchez of U.S. Immigration and Customs Enforcement’s (ICE) Office of Principal Legal Advisor pleaded guilty today for a wire fraud and aggravated identity theft scheme involving the identities of numerous aliens . . . .”). As a result, the concerns of the eleven million undocumented persons within U.S. borders find jus­tification in both past practice and the current Administration’s words and actions. 145 As the judge in the New York challenge noted, “[A] higher ‘level of concern about using citizenship data for enforcement purposes[ ]’ could also exacerbate the effects of adding a citizenship question.” New York v. U.S. Dep’t of Commerce, 351 F. Supp. 3d 502, 579 (S.D.N.Y. 2019), cert. granted, 139 S. Ct. 953 (2019) (mem.) (quoting J. David Brown et al., Understanding the Quality of Alternative Citizenship Data Sources for the 2020 Census 39 (2018), []).

If the Supreme Court is ultimately unpersuaded by these arguments and the citizenship question is included on the 2020 short form, however, it will be even more imperative to challenge other threats to an accurate count of communities of color, such as the Census’s criteria for counting incarcerated persons. 146 See supra notes 17–20 and accompanying text; infra section III.B.

B. Prison-based Gerrymandering

Prison-based gerrymandering results from the exportation of black and brown bodies to correctional facilities in white rural communities and the Census’s counting incarcerated persons in their place of impris­onment instead of the community of their last place of residence. This practice, in turn, results in the siphoning of both political and financial resources away from underfunded neighborhoods. 147 See, e.g., Grace Dixon, How Prison Gerrymandering Strips Power from Communities of Color, In These Times (Aug. 16, 2018), [] (citing American Community Survey data that show that rural communi­ties—which house 40% of the nation’s predominantly black and Latinx prison population and count these individuals as residents—are 79% white, while urban areas are 56% people of color). This mutually rein­forcing pattern of resource deprivation and incar­ceration is fueled, in part, by manipulating the enumeration of prison popu­lations—which are disproportionately African American and Latinx—to the benefit of white rural communities and by undercounting communities of color. 148 See id. (“Prison-based gerrymandering distorts representation within states by counting prisoners as residents of the location where they are incarcerated when drawing state district maps, rather than at their permanent residences. This imbalance results in a simultaneous transfer of both residents and political power, disproportionately denying minor­ity communities equal representation.”). Thus, the district in which incarcerated persons are counted for Census pur­poses reflects a policy choice that has a disproportionate impact on black and brown communities.

The impact of that policy choice is further amplified by the sheer size of the U.S. prison population. The United States’ project of mass incar­ceration has distinguished it as the country with the highest incarcera­tion rate in the world. 149 See Highest to Lowest—Prison Population Rate, World Prison Brief, [] (last visited Feb. 2, 2019). It is well understood that American prisons and jails are disproportionately populated by black and brown communities. The collateral consequences of such a deeply racially skewed carceral state, however, are often overlooked. One such consequence is the “miscount­ing” of most of the country’s 2.3 million incarcerated persons 150 Peter Wagner & Wendy Sawyer, Mass Incarceration: The Whole Pie 2018, Prison Policy Initiative (Mar. 14, 2018), [] (compiling the total number of incarcerated persons in America across all jurisdictions and systems of confinement). in the decen­nial census. Forty percent of the people in the nation’s prisons and jails are black, which is more than three times the representation of black people in the general population. 151 Id. As of 2016, 23% of the prison pop­ulation was Latinx, 7% greater than the percentage of Latinx persons in the general adult population. 152 See John Gramlich, The Gap Between the Number of Blacks and Whites in Prison Is Shrinking, Pew Research Ctr.: Fact Tank (Jan. 12, 2018), [] (“[W]hile Hispanics represented 16% of the adult popu­lation, they accounted for 23% of inmates.”).

The location of correctional facilities also matters. As one scholar observed, “Correctional facilities are not dispersed evenly throughout most states, but are often found in more rural, predominantly white areas, while people incarcerated in these facilities are disproportionately people of color from comparatively urban areas.” 153 Ebenstein, supra note 32, at 325. As a result, commu­nities that contain such facilities stand to gain from the imported body count that incarceration produces. The more population a jurisdiction can claim, the more representation it is entitled to, regardless of whether those individuals are represented in the political process. This phenom­enon is part of a historical continuum of counting significant numbers of people of color but not according them representational equality. From enslavement to Jim Crow to mass incarceration coupled with felon-disenfranchisement laws, black and brown citizens and noncitizens alike have helped to swell population numbers in jurisdictions where they remain politically invisible or impotent.

Put simply, in the context of prison-based gerrymandering, black and brown incarcerated bodies increase the population count of rural, mostly white communities where, with few exceptions, 154 Incarcerated persons are permitted to vote in Maine and Vermont. See Me. Rev. Stat. Ann. tit. 21-A, § 111 (2017) (containing no restriction on the right to vote based on criminal status); Vt. Stat. Ann. tit. 17, § 2121 (2018) (same). they do not vote and therefore have no means of holding local governments and repre­sentatives accountable. At the same time, those bodies are removed from the population count of the communities in which they would otherwise reside but for their incarceration, leaving those communities to which they are most likely to return, and in which their relatives are more likely to reside, with a deflated population count. According to Justice Ginsburg’s definition of “representational equality” in Evenwel, 155 See supra notes 22, 90–95 and accompanying text. counting prisoners in communities in which they have no ties reduces the likelihood that their interests will be represented there.

Despite robust advocacy from civil rights groups, including LDF, the Census Bureau has refused to update its policy of counting incarcerated persons as residents of their detention facilities. 156 See, e.g., Letter from Sherrilyn A. Ifill, President & Dir. Counsel, & Leah C. Aden, Assistant Counsel, NAACP LDF, to Karen Humes, Chief, Population Div., U.S. Census Bureau (July 19, 2015), [] (“Beginning with the 2020 Census and each subsequent decennial census, LDF urges the Census Bureau to count incarcerated people as residents of their last known pre-arrest home address, rather than of the particular prison facility where they happen to be located on Census day.”); Comments on 2020 Census Residence Rule and Residence Situation, The Leadership Conference on Civil & Human Rights (Sept. 1, 2016), []; LDF Disappointed in the U.S. Census Bureau’s Proposed Residence Criteria for Incarcerated Persons for 2020 Census, NAACP LDF (July 7, 2016), []. Like other temporal relocations such as certain military service, stays in health­care facilities, and college attend­ance, incarceration does not establish a durable connection to a commu­nity in which incarcerated individuals ordinarily engage as full members of society. Currently, the Census Bureau defines “usual residence” as the place where one “lives and sleeps most of the time.” 157 Final 2020 Census Residence Criteria and Residence Situations, 83 Fed. Reg. 5525, 5526 (Feb. 8, 2018) (to be codified at 15 C.F.R. ch. 1). Not only does this narrow definition often result in the extraction of fiscal resources from many black and Latinx communities to rural, mostly white areas that host correctional facilities, the artificial deflation and inflation of the populations in these respective communities can also severely impact political representation when these data are used for redistricting purposes.

New York, California, Maryland, Delaware, and over 200 local juris­dictions mandate that incarcerated individuals be counted as resi­dents of their home communities for redistricting purposes. 158 See Ebenstein, supra note 32, at 360; see also Prison Policy Initiative, Solutions, Prison Gerrymandering Project,
[] [hereinafter PPI, Solutions] (last visited Jan. 29, 2019) (noting that “more than 200 rural governments have implemented” solutions to count incarcerated people as residents of their home); supra note 32 and accompanying text.
However, the vast majority of jurisdictions across the country rely on skewed Census data to allocate population pursuant to the one-person, one-vote mandate for congressional, state, and local districting. While there are legal chal­lenges that seek to expand fair counting practices to other states, 159 See, e.g., Mary E. O’Leary, NAACP, Yale Law Sue State over ‘Prison Gerrymandering,’ New Haven Reg. (June 28, 2018), [] (discussing the NAACP and Yale Law School’s suit alleging that prison gerry­mandering violates the one-person, one-vote principle of the Equal Protection Clause). changing the Census Residence Criteria would provide the most sweeping change to rebalance the fraught and discriminatory population attribution that these criteria create.

For example, the 2000 Census counted only 17% of Maryland’s incarcerated population in Baltimore 160 Garima Malhotra, Maryland Ends Prison-Based Gerrymandering, Brennan Ctr. for Justice (May 4, 2010), []. even though the city produced 68% of the state’s total population of incarcerated people. 161 Prison Policy Initiative, Fixing Prison-Based Gerrymandering After the 2010 Census: Maryland, Prison Gerrymandering Project (Mar. 2010), []. As a result, “[m]uch of the population gains of Western Maryland and the Eastern Shore from these prison populations have been at the expense of Baltimore City, which has been losing representation in the State Legislature since the 1980’s.” 162 Michelle Davis, Assessing the Constitutionality of Adjusting Prisoner Census Data in Congressional Redistricting: Maryland’s Test Case, 43 U. Balt. L.F. 35, 37 (2012). Maryland has since passed the No Representation Without Population Act of 2010, which mandates the use of adjusted population data that count incarcerated persons at their last-known residence for purposes of congressional, state, and local redistrict­ing. 163 See No Representation Without Population Act, ch. 67, 2010 Md. Laws 737 (codified at Md. Code Ann., State Gov’t §2-2A-01 (West 2018)) (“The population count . . . shall count individuals incarcerated in the state or federal correctional facilities, as determined by the decennial Census, at their last known residence before incarceration if the individuals were residents of the state.”). New York adopted comparable legislation in the face of similarly skewed redistrict­ing outcomes between its upstate population, where most of its prisons are located, and its downstate population, from which most of its incarcerated population comes. 164 See Act of Aug. 11, 2010, ch. 57, pt. XX, 2010 N.Y. Laws 315, 405 (codified at N.Y. Correct. Law § 71 (McKinney 2018)). For example, seven state senate districts met mini­mum population requirements only because the Census counted persons incarcer­ated upstate as residents there. Prison Policy Initiative, Gerrymandering and Relying on the Miscount of Prisoners Combine to Violate the U.S. Constitution in New York, Prison Gerrymandering Project, [] (last visited Jan. 30, 2019). Before enacting a legislative remedy, “[p]rison-based gerrymandering helped the New York State Senate add an extra district in the upstate region. Without using prison populations as padding, seven state senate districts would have to be redrawn, causing line changes throughout the state.” 165 Prison Policy Initiative, The Problem, Prison Gerrymandering Project, [] [hereinafter PPI, The Problem] (last visited Jan. 29, 2019).

New York responded to the artificial inflation of these legislative districts by passing legislation to adjust the population data after the 2010 Census to count incarcerated people at their respective pre-incarceration homes for state and local redistricting purposes. 166 Act of Aug. 11, 2010, ch. 57, pt. XX, § 1. The statute provides that “[i]n each year in which the federal decennial census is taken but . . . does not implement a policy of reporting incarcerated persons at each such person’s residential address prior to incar­ceration, the department of correctional services shall . . . deliver to the legislative task force on demographic research and reapportionment” information for each incarcerated person that identifies his or her pre-incarceration residence and location of incarceration when the Census was taken. Id. This information is then used to adjust Census data for purposes of state senate, state assembly, county, and municipal districting. Id. While this more equitably distributes political resources, it does not affect federal funding based on Census data. In other jurisdictions, massive imbal­ances persist, resulting in palpable dilution of the minority vote. For example, Cook County, Illinois, where Chicago is located, pro­duces 60% of the state’s incarcerated population, 167 PPI, The Problem, supra note 166. which is 55.9% black and 12.8% Latinx. 168 Ill. Dep’t of Corr., Adult Inmate Population on June 30, 2018 (2018), []. The 2010 Census counted 99% of that population outside of Cook County, 169 PPI, The Problem, supra note 166. however, diminishing the political power of the Cook County residents who are 24.8% black and 24% Latinx. 170 American Fact Finder, U.S. Census Bureau, [] (last visited Feb. 21, 2019). Like­wise, in Texas, there is a state legislative district in a rural area composed of 12% incarcerated persons, which means that residents from that district are overrepresented in the state legislature vis-à-vis denser, urban areas in the state. 171 Prison Policy Initiative, Fixing Prison-Based Gerrymandering After the 2010 Census: Texas, Prison Gerrymandering Project,
50states/TX.html [] (last visited Feb. 3, 2019).

To mitigate this distortion, the Census Bureau has stated that, following the 2020 Census, it plans to offer a product to states that will enable them to reallocate their incarcerated population to their pre-incarceration addresses. 172 See Final 2020 Census Residence Criteria and Residence Situations, 83 Fed. Reg. 5525, 5527–28 (Feb. 8, 2018) (to be codified at 15 C.F.R. ch. 1). The Census will require the state to provide the pre-incarceration address and the location of incarceration on Census day in order to carry out this change. See id. Although this would allow states to rely on more accurate data for purposes of state and local redistricting and resource allocation, it will neither affect the official decennial census count 173 See id. nor ensure compliance with the principle of one-person, one-vote and the Voting Rights Act’s protection of minority communities’ vot­ing strength. Specifically, this product is an inadequate antidote to under­count because, while it enables states to count incarcerated persons at their pre-incarceration residence, it does not reverse the miscount on the actual Census which is used for federal resource allocation and political repre­sentation at the federal level.

Similarly, following the Census, the Bureau publishes data called the “Advance Group Quarters Summary File” that show which areas contain correctional facilities. 174 Id. States can use this information to exclude the population of correctional facilities from a subdivision’s total population for state and local redistricting. 175 Id. While the Census Bureau has indicated that it will provide such “group quarters data” for the 2020 Census sooner than it did following the 2010 Census, 176 Id. “helping data users remove prison populations counted at the prison location does not solve the parallel problem that the Census Bureau failed to count these popula­tions in the correct place.” 177 PPI, Solutions, supra note 159.

Ultimately, an updated and more accurate rule that counts incarcerated people at their pre-arrest address, rather than at the prison facilities where they are incarcerated for federal purposes, is necessary to ensure a more robust democracy for the benefit of our country’s resi­dents. In addition, there is active litigation challenging prison-based gerry­mander­ing in Connecticut 178 Associated Press, Racial, Prison Gerrymandering Lawsuits to Make for a Busy Year of Legal Battles as Redistricting Nears, Atlanta Black Star (Jan. 15, 2019), [] (citing prison gerrymandering suits in Connecticut). and other longstanding advocacy efforts underway to challenge prison-based gerrymandering that, if suc­cessful, can help mitigate the Census’s present inaccuracies and inequi­ties on a piecemeal basis.

B. Displacement, Downturn, and Natural Disaster

Climate-related displacement and the ongoing effects of the 2008 eco­nomic downturn, which caused a disproportionate loss of home ownership and displacement among people of color, 179 Sarah Burd-Sharps & Rebecca Rasch, Soc. Sci. Research Council, Impact of the US Housing Crisis on the Racial Wealth Gap Across Generations 3, 19 (2015), [] (“By 2031, white wealth is forecast to be 31 percent below what it would have been without the Great Recession, while black wealth is down almost 40 percent. For a typical black family, median wealth . . . will be almost $98,000 lower than it would have been without the Great Recession.”); Gillian B. White, The Recession’s Racial Slant, Atlantic (June 24, 2015), [] (“Wealth often determines not only how well families can provide for themselves when it comes to basics like food and shelter, but it is a safety net for emer­gencies and helps to set up future generations for education, home owner­ship, and other opportunities that improve people’s lives.”). have combined to exacerbate the undercount. Communities of color are disproportionately impacted by natural disasters and economic downturns, both of which fuel displacement and alter population growth patterns. 180 See Christian E. Weller, Jaryn Fields & Folayemi Agbede, Ctr. for Am. Progress, The State of Communities of Color in the U.S. Economy 1 (2011), https://cdn.american [] (“All U.S. households were severely hurt by the recession but communities of color experienced larger losses than whites. This also means that . . . communities of color will have to climb out of a deeper hole to regain the same level of economic security as they had before the crisis.”); Junia Howell & James R. Elliott, Damages Done: The Longitudinal Impacts of Natural Hazards on Wealth Inequality in the United States, Soc. Probs. (forthcoming) (manuscript at 3), (on file with the Columbia Law Review) (“Results from longitudinal models indicate that as local hazard damages increase, so too does wealth inequality, especially along the lines of race, education, and homeownership. Results also indicate that the more funds areas receive from FEMA, the more this wealth inequality increases.”).
As compared to whites, the financial stability of communities of color is more precari­ous, making them more susceptible to deracination and dispersion when calamity strikes, which may result in the omission of their house­holds in Census enumeration. Following Hurricane Katrina, for example, the population loss in New Orleans resulted in the loss of a congressional seat for the state of Louisiana and a population increase in cities and towns of nearby states, such as Houston, Texas, which took in the greatest number of migrants of the post-Katrina diaspora. 181 See Kimball W. Brace, Election Data Servs., Katrina-Driven Population Loss Will Also Lead to Louisiana’s Loss of a Congressional Seat, According to 2006 Population Estimates 1 (2006), []; Tom Benning, Texas’ Katrina Population May Be Key to State Gaining a 4th House Seat, Dall. Morning News (Jan. 20, 2010), [].

In the past two years alone, severe storms—including Hurricane Irma, which impacted Miami and the Florida Gulf Coast, and Hurricane Maria, which wreaked havoc on Puerto Rico—created an influx of migration to Florida that may impact whether it gains an additional congressional seat. 182 See Martín Echenique & Luis Melgar, Mapping Puerto Rico’s Hurricane Migration with Mobile Phone Data, CityLab (May 11, 2018), [https://] (“Most of those who left the island first moved to Florida (43 percent) . . . .”).
In 2018, Hurricane Michael devastated areas of Florida. It is possible that one of the reverberating consequences of these natural disasters will be the exacerbation in undercount or displacement of communities of color.

Displacement is also a product of economic downturns, the effects of which impact the states whose populations most suffer the fallout from these downturns. For example, “states with high foreclosure rates (Arizona, Nevada, Florida, and Michigan) could be adversely affected if doubled-up households were left out of the count.” 183 Expect the Unexpected: Apportionment Projections Based on Census Bureau Estimates, Mich. Population Studies Ctr., Inst. for Soc. Research, [] [hereinafter Expect the Unexpected] (last visited Jan. 30, 2019). “Doubled-up households” refer to the occupancy of “one or more adults in addition to the head of household and spouse or partner, such as an adult child living at home, two related or unrelated families residing together, or a parent living with an adult child” in a single household. 184 American Housing Survey Reveals Rise in Doubled-Up Households During Recession, PD&R Edge, HUD User,
research_012714.html [] [hereinafter American Housing Survey] (last visited Jan. 30, 2019) (“[F]rom 2003 to 2009, the rate of new household formation declined and the number of ‘doubled-up’ households increased.”).
As of 2009, just before the last decennial census, “[b]lack, Hispanic, and other nonwhite households were more likely to be doubled up than were white households, and households with a foreign-born head were more likely to be doubled compared to those with a native-born head.” 185 Id. Unsurprisingly, doubled-up households are harder to count, 186 See Expect the Unexpected, supra note 184 (“It is more likely that the [Community Population Survey] will pick up doubled-up families than the self-enumerated census will.”). and they increased substantially as a result of the 2008 economic recession. 187 See American Housing Survey, supra note 185 (“[T]he recent financial crisis and subsequent recession influenced the rise in doubled-up households.”); David Johnson, Households Doubling Up, Census Blogs, U.S. Census Bureau (Sept. 13, 2011), [] (reporting an increase in doubled-up households from 19.7 million in 2007 to 21.8 million in 2011). A 2010 Current Population Survey (CPS) revealed “an 11.6 percent increase in the number of dou­bled up families.” 188 Expect the Unexpected, supra note 184. These factors suggest that most economically fragile communities, especially communities of color, will continue to be under­counted without an appropriate adjustment of Census data.

D. Digital Divide

Finally, the expanded use of the internet to administer the Census presents a counterintuitive dilemma. Over time, Census respondents have increas­ingly declined to respond to Census forms sent by mail. 189 See Nat’l Research Council, Modernizing the U.S. Census 44–46 (Barry Edmonston & Charles Schultze eds., 1995) (“If mail response rates continue to decline, as they have done for the past several censuses, the national mail response rate may be less than 60 per­cent in 2000.”). Now, most U.S. households (approximately 55%) that would ordinarily have received the short form in the mail will receive a unique security identifier in the mail that they can use to fill out a Census survey online. 190 D’Vera Cohn, For 2020, Census Bureau Plans to Trade Paper Responses for Digital Ones, Pew Research Ctr.: Fact Tank (Feb. 24, 2016), []. People will also be permitted to call in and answer the ten ques­tions on the Census form. 191 Alexis Farmer, Digitizing the 2020 Census, Brennan Ctr. for Justice, [] (last visited Jan. 30, 2019). While online administration of a digitized Census would seemingly increase accessibility, there are concerns about the security and equity of this process.

Communities of color and poor, rural, and immigrant communities, all of whom historically have been undercounted, are all less likely to have broadband access. A Pew Research Center survey using 2016 data revealed that 34% of African Americans and 40% of Latinx persons do not own a desktop or laptop computer, while 35% of African Americans and 42% of Latinx persons do not have broadband internet at home. 192 Andrew Perrin, Smartphones Help Blacks, Hispanics Bridge Some—But Not All—Digital Gaps with Whites, Pew Research Ctr.: Fact Tank (Aug. 31, 2017), []; see also Sam Gustin, Systemic Racial Discrimination Worsens the US Digital Divide, Study Says, Motherboard (Dec. 14, 2016), [] (“While 81 percent of Whites and 83 percent of Asians have home inter­net . . . only 70 percent of Hispanics, 68 percent of Blacks, 72 percent of American Indian/Alaska Natives, and 68 percent of Native Hawaiian/Pacific Islanders are connected at home.”). The same Pew Research Center data indicated that approximately one-third of Americans making less than $30,000 a year do not own a smartphone. 193 Monica Anderson, Digital Divide Persists Even as Lower-Income Americans Make Gains in Tech Adoption, Pew Research Ctr.: Fact Tank (Mar. 22, 2017), []. Often overlooked are people with limited English profi­ciency, only 51% of whom have internet access, and people with disabilities, only 50% of whom reported daily internet use as compared to 79% of persons without disabilities. 194 The Leadership Conference Educ. Fund, Counting Everyone in the Digital Age 10 (2017), []. Data from the 2015 National Content Test and the American Community Survey suggest that using a mobile device to respond to questions was “more burdensome and likely produced less accurate data than using a desktop or laptop computer. Mobile device users faced the challenges of longer load time for the ACS web application, small font sizes, using a finger to select the proper option, and more scrolling.” Id. at 14 (footnote omitted). These problems will be difficult to rem­edy without adequate testing. In addition, people living in the South—home to over half the African American population, 195 Regional Distribution of the Black Population, Black Demographics, [] (last visited Apr. 10, 2019) (citing that more than fifty percent of the U.S. black popu­lation resides in the South). as well as the greatest number of states with low rates of high-speed internet and people living in rural areas—face additional challenges to digital accessibility. 196 See The Leadership Conference Educ. Fund, supra note 195, at 10 (noting that the South is the least connected region in terms of Internet access and that “most states with especially low rates of high-speed Internet were in the South”).

In order to prevent hacking, especially in the face of several breaches of high-profile, high-user, well-resourced platforms, 197 See Taylor Armerding, The 18 Biggest Data Breaches of the 21st Century, CSO (Dec. 20, 2018), [] (citing breaches to platforms such as Yahoo, eBay, and others). and to trouble­shoot the system before the Census rollout, the government had sched­uled tests of the program. 198 2017 Census Test, U.S. Census Bureau, [] (last updated May 12, 2017). However, the Census Bureau scuttled three of these tests­—one in Washington State near a Native American reserva­tion, one in the Standing Rock Reservation in North and South Dakota, and another in Puerto Rico—due to lack of funding. 199 John H. Thompson, U.S. Census Bureau Announces Changes to 2017 Field Tests, U.S. Census Bureau (Oct. 18, 2016), []. While tech­nology can bring greater efficiencies and access to a variety of gov­ern­ment functions, to yield the greatest good, it must be supported with adequate testing, appropriate security, and efforts to ensure equity among histori­cally disenfranchised groups. With respect to the Census, the technology that could potentially enable greater access to the Census may become a disproportionate liability for communities of color that find themselves on the wrong side of the digital divide in an undertested and insecure Census apparatus.

IV. Prescriptions to Improve Population Count

On May 8, 2018, the U.S. Government Accountability Office released the testimony of its Director of Strategic Issues and its Director of Infor­mation Technology in which they identified ways to mitigate the primary risks jeopardizing a cost-effective and secure enumeration. 200 See GAO 2020 Census Report, supra note 112, at 1–5 (noting that, despite the difficulty in “achieving a complete count” due to “external societal challenges” and “inter­nal management challenges,” the Census Bureau made eighty-four recommendations to help address the numerous risks and issues presented). These interventions—including comprehensive testing of IT and method­olog­ical innovations, improved security for Census systems, and increased funding and oversight—are critical to a functional Census. 201 See id. at 6–21 (highlighting a variety of key risks to the administration of the Census and discussing four broad innovation areas to meet these risks: “re-engineering field operations, using administrative records, verifying addresses in-office, and developing an Internet self-response option”). Given the current Administration’s aggressive embrace of policy choices that under­mine the constitutional mandate to enumerate all persons within U.S. borders, however, it seems unlikely the federal government will pursue these suggested solutions.

In the face of the federal government’s abdication of responsibility, local actors and civil society can play an important role in promoting a fair and accurate enumeration of the U.S population. Knowing the dis­crim­inatory history of the Census and understanding the current stakes should embolden advocates to continue to aggressively oppose any mea­sures that threaten to exacerbate undercount, especially the inclu­sion of the citizenship question, and empower communities to demand greater protections and accountability from the Census Bureau. This Part explores possible interventions by state and local public and private stakeholders to increase the integrity of the 2020 Census and offers a roadmap to course correct in advance of the most important tally we per­form as a nation.

A. Use Interest Convergence to Broker Power on Behalf of the Emerging Majority

State and local governments, businesses, communities, and individ­uals all benefit when residents are fully counted. Census data provide information that can be used as currency for planning, politics, and profit. When the interests in gathering this information converge with the interest of undercounted populations to be accurately enumerated, there are opportunities for cooperation and collaboration to produce a better Census. Such interest convergence 202 Interest convergence theory was first defined by the late Dr. Derrick Bell, former LDF attorney and renowned law professor. He defined interest convergence as a circum­stance when interracial allegiances can be formed to achieve a mutually beneficial outcome. See Derrick A. Bell, Jr., Brown v. Board of Education and the Interest-Convergence Dilemma, 93 Harv. L. Rev. 518, 523–24 (1980) (describing the principle of interest convergence). is particularly useful when there is a threat to participation in the Census that will deprive everyone of essential data about the American populace.

In addition to the inherent frailties of an underfunded and inade­quately tested and supervised Census, there are many understandable concerns and fears about the potentially harmful uses of Census data that threaten to quell participation. Despite victories in all three pending cases challenging inclusion of the citizenship ques­tion, the threat still looms until the legal battle is fully resolved in favor of its exclusion. 203 This Essay will go to print shortly before the Supreme Court is expected to decide the issue of whether the citizenship question can be included in the Census short form. Regardless of the outcome, these prescriptions can be used to enhance the political power of the emerg­ing majority populations, but they will be especially needed if the question is included. In order to be both fully counted and in a position to hold those in power to account, the emerging majority must vigorously engage in the enu­meration process. Transparency concerning the current and poten­tial use of Census data is crucial. Under current law, no personally identifying information can be lawfully furnished to another government agency without consent of the subject of the information for seventy-two years from the time such information is acquired by the Census. 204 Pursuant to the “72-Year Rule,” the U.S. government is prohibited from releasing personally identifiable information about an individual to any other individual or agency until seventy-two years after it was collected for the decennial census, at which time the rec­ords will be released to the public by the National Archives and Records Administration. Act to Amend Chapter 21 of Title 44, Pub. L. No. 95-416, 92 Stat. 915 (1978); see also History: The “72-Year Rule,” U.S. Census Bureau, [] (last visited Mar. 7, 2019). Before such time, access to personally identifiable infor­mation from the Census is restricted to the individual named on the record or their legal heir. Id. And yet, given the current Administration’s hostility toward certain immigrant populations, it is both impossible and imprudent to offer minority com­mu­nities with undocumented residents unqualified assurances that engaging the Census is without any risk. Instead, we must strike a balance between fueling enumeration and minimizing exposure.

The sobering reality is that the numerical power of the emerging majority will not necessarily translate into fiscal or political power if sizeable portions of those communities remain invisible. Therefore, commu­nities of color cannot afford to be complicit in creating false evidence of nonexistence by opting out of the Census. We must make careful judgments for ourselves and our families about how to answer Census questions with the ultimate goal of finding a way to be counted.

Public education begins with translating to the most marginalized communities, their neighboring communities, and the elected officials of their states what they all stand to gain by both participating in the Census and protecting the safety of all residents. This translation may take the form of elected officials taking action in advance of the Census to become a “sanctuary city” 205 The term “sanctuary cities” or “safe cities” refers to jurisdictions that have policies that limit the role of and cooperation with local, state, and federal immigration authorities in the enforcement of federal immigration laws. See, e.g., N.Y. State Att’y Gen., Guidance Concerning Local Authority Participation in Immigration Enforcement and Model Sanctuary Provisions 1 n.1 (2017), [https://]; Immigration 101: What Is a Sanctuary City?, America’s Voice (Apr. 25, 2017), [] (last updated Oct. 9, 2018).
—making express the civil and human rights protections in state and local laws and articulating a message of inclusion followed by concrete action that recognizes the important influence of various ethnic and racial groups within a jurisdiction. For example, Director of U.S. Immigration and Customs Enforcement (ICE) Sarah Saldaña testified before Congress in 2015 that “over 200 jurisdictions, including some of the largest in the country, . . . refuse to honor ICE detainers, while some have also denied ICE access to their jails and prisons.” 206 A Review of the Department of Homeland Security’s Policies and Procedures for the Apprehension, Detention, and Release of Non-Citizens Unlawfully Present in the United States (Part II): Hearing Before the H. Comm. on Oversight and Gov’t Reform, 114th Cong. 12 (2015) (statement of Sarah Saldaña, Director, U.S. Immigration & Customs Enforcement). In addition, other cities have provided services for immigrant populations that aim to acclimatize immigrants and supply the resources needed to function with greater ease in society, such as “offering English language classes; issuing municipal identification documents and driver’s licenses to all residents; ensuring  that  immigrants  have  equal  access  to  bail; establishing  U-Visa  policies; and  training  criminal  prosecutors  and  public  defenders on  the  immigration  consequences of convictions and plea deals.” 207 Am. Immigration Council, “Sanctuary” Policies: An Overview 1 (2017), https:// [].
These practices and policies signal an interest in protect­ing and providing for immigrant communities which could, in turn, encourage greater Census participation. These measures are recom­mended even if the citizenship question is not included on the Census because they will bolster participation and serve as a partial antidote to the fear created by the threat of the question.

Even if emerging minority populations are fully counted in the Census, it is not guaranteed that those numbers will translate into adequate political power for purposes of redistricting and funding alloca­tion. Indeed, systemic discrimination, voting-status limitations, and active partisan and racial gerrymandering are among several obstacles that complicate the political empowerment of the emerging majority. Many of the states that stand to gain population by virtue of the substan­tial growth of their minority communities are also serial violators of voting rights protections and active culprits of voter suppression. 208 For example, in North Carolina, while the black population of 2.2 million constitutes 22% of the state’s total population, Jessica Stanford, NC in Focus: Black Population in North Carolina, 2016, Carolina Demography (Feb. 8, 2018), https:// [], and the Latinx population is approach­ing one million, having increased from just over 75,000 in 1990 to 800,000 in 2010, Rebecca Tippett & Krista M. Perreira, The Hispanic/Latino Community in North Carolina, Carolina Demography (Oct. 10, 2017),
hispaniclatino-community-in-north-carolina/ [], there have been multiple schemes this past decade designed to disenfranchise Black voters with “surgical precision,” NAACP v. McCrory, 831 F.3d 204, 214 (4th Cir. 2016).
For these reasons, preconditions for Census participation should be negoti­ated and set in advance. As “all politics are local,” so are the specific power brokers necessary to be at the negotiating table. Depending on where the emerging majority growth is, communities may lobby mayors, city councils, state legislatures, or governors to provide assurances that political representation will be based on population growth within the bounds of the Constitution, the Voting Rights Act, and state and local laws. For example, if a state stands to gain a congressional seat as a result of a substantial increase of its Latinx population in a specific region, communities might demand that the state legislature commit to increas­ing representation in that region, assuming that the Census numbers produce this result.

The interest convergence between state and local government officials and their minority communities to ensure an accurate enumera­tion is the linchpin of securing adequate federal funding and accurate state-level funding allocation. Without adequate resources to care for their resident populations, a climate of social division and competition for limited resources may envelop state and local governments, which would in turn adversely affect the economic and social health of the state. By embracing the nation’s diversity as one of their most powerful assets, states can compete for dollars and influence on behalf of their respective emerging majority populations that, counted or not, will not disappear.

B. Leverage Voter Turnout Efforts for the 2018 Midterm Elections to Create a Census Mobilization Apparatus

The 2020 Census is coming on the heels of what was arguably one of the most consequential U.S. midterm elections in recent history. 209 Eric Lach, How This Election Compares to Past Consequential Midterms, New Yorker: The Current (Nov. 6, 2018), (on file with the Columbia Law Review) (noting that “some midterms shape more than others” and that the November 2018 midterms have been compared “to those that took place in 1974, after Richard Nixon’s resignation; in 1994, when Newt Gingrich led the ‘Republican Revolution’; and in 2010, when the Tea Party swept many Republicans to power”); Kenneth Lowande, Melinda Ritchie & Erinn Lauterbach, Having the Most Diverse Congress Ever Will Affect More than Just Legislation, Wash. Post (Jan. 9, 2019), (on file with the Columbia Law Review) (suggesting that the record number of elected officials of color and veterans will result in more advocacy on behalf of those affinity groups). Voter turnout was higher in this midterm election than it has been in any other midterm election in over a century. 210 See, e.g., Ed Kilgore, 2018 Turnout Was the Highest of Any Midterm in More than a Century, N.Y. Mag.: Intelligencer (Nov. 13, 2018), []; Niall McCarthy, Midterm Elections See Highest Turnout in More than a Century, Forbes (Nov. 16, 2018), []. The midterms ushered in the most racially diverse Congress ever, along with a historic number of women and several other firsts, including the first Native American and Muslim congresswomen 211 See, e.g., Kimberly Truong, The New Face of Power Officially Takes Control Today, N.Y. Mag.: The Cut (Jan. 3, 2019), []. and the first openly gay male governor. 212 Michael Stratford, Colorado Elects Nation’s First Openly Gay Governor, Politico (Nov. 6, 2018), []. The enthusi­asm and engagement that brought voters to the polls in November 2018 present an opportunity to galvanize the voters who sup­port­ed these and other candidates to serve as foot soldiers for the Census. 213 See Christopher Parker & Henry Fernandez, There’s a Boost in Black Turnout, Especially Among Black Women Voters, Hill (Aug. 28, 2018), [] (describing the effects of black voter turnout on progres­sive victories in Alabama, Georgia, and Virginia). Even if they are unable to vote, community members can be encouraged to be counted in the Census as a means of civic engagement. Census participation can then be leveraged as a gateway to broader civic participation and organ­ization regardless of whether one is currently eligible to vote.

Indeed, in order to gain and maintain any political power transmit­ted through population count and apportionment, it is essential that marginalized communities engage not just on Election Day but year-round. It is critical to message the connection between voting rights, representation, and an accurate Census. For example, certain election infrastructure is tied to federal funding under the Help America Vote Act of 2002 (HAVA). 214 42 U.S.C. §§ 15301–15545 (2012). The appropriations bill for the 2018 elections provided $380 million to the Election Assistance Commission to distribute to states for purposes of improving federal election administration, including investments in technol­ogy and security. See 2018 HAVA Election Security Funds, U.S. Election Assistance Comm’n, [] (last visited Jan. 30, 2019). HAVA uses a Census derivative voting-age-population formula to allocate funds to states through the U.S. Election Assistance Commission for a range of election administration activities, from up­grad­ing electronic voting systems to eliminate vulnerabilities to cyber­attack to funding acquisition of voting machines with a verified paper trail. 215 42 U.S.C. §§ 15301–15303. Thus, the population count of each state informs the amount of funding needed to administer elections. Similarly, the allocation of election equipment, polling sites, and poll workers is tied to population count. 216 See EAVS Deep Dive: Poll Workers and Polling Places, U.S. Election Assistance Comm’n (Nov. 15, 2017), [] (“2016 data demonstrated wide variation among and within the states on [registered voters per polling place]. For example, . . . 25 percent of reporting jurisdictions had an average of more than 2,000 registered voters per polling place, 27 percent had between 1,000 and 2,000, and 48 percent had less than 1,000.”). In these basic but substantial ways, ensuring that communities of color are counted can enable them to exercise greater political power at the ballot box.

C. Pass Omnibus Legislation Creating Nonpartisan Redistricting Commissions, Mandating Sampling, and Implementing a “Usual Residence” Rule for Incarcerated Persons

States that stand to gain increased congressional representation be­cause of a disproportionate increase in their minority populations should pledge to pursue a distribution of political power that acknowl­edges that numerical leverage. A first step would be to establish a nonpartisan redistricting commission with implementing legislation that requires the use of sampling for redistricting and that counts incar­cerated persons at their pre-incarceration residence. More  progressive  measures  would  include  a  commitment  to  noncitizen  voting  in  certain  state  and  local  elec­tions to  demonstrate  a  willingness  to  acknowledge  a  diversity  of  voices  in  local governance. 217 Many other countries, including in the European Union, permit noncitizen voting in their elections. See Ron Hayduk, Why Non-Citizens Should Be Allowed to Vote, Jacobin (Nov. 6, 2018), [] (“Globally, more than forty-five countries on nearly every continent permit voting by resident immigrants.”). Domesti­cally, noncitizen voting already exists in several municipalities across the United States. See id. (“Today, noncitizen voting is most common in Maryland: ten municipalities allow it in all local elections . . . . In 2016, the same day Trump was elected, San Franciscans passed a ballot measure opening the franchise to noncitizen parents.”). In addition, changing the Census’s Residence Criteria to count incarcerated persons at their pre-incarceration address would yield a significant population redistribution that more accurately reflects the familial and communal ties of incarcerated persons and those whose fates are most closely linked to theirs.

Finally, when Census data fall short on accuracy, sampling—which involves surveying a subset of the population and extrapolating those data to make observations about and attribute characteristics to the whole population—can be used to decrease the disparity in undercount, more equitably allocate federal funding, and increase fairness in redistricting. Sampling arose out of the aftermath of the 1990 Census, which con­tained a staggering undercount of four million people, including 4.4% of African Americans, 5% of Hispanics, and 12.2% of Native Americans. 218 Howard Hogan & Gregg Robinson, What the Census Bureau’s Coverage Evaluation Programs Tell Us About Differential Undercount, in U.S. Census Bureau, 1993 Research Conference on Undercounted Ethnic Populations: Proceedings 9, 18 tbl.3 (1993).

A Republican-led legal challenge to the use of sampling in congressional districting led to its ban for that purpose in a 5-4 decision by the Supreme Court in 1999. 219 Dep’t of Commerce v. U.S. House of Representatives, 525 U.S. 316, 318, 343 (1999) (holding 5-4 that statistical sampling is inconsistent with the Census Act). However, “sam­pling” or representative data are successfully used in many other contexts. See, e.g., Lesson 4: Use of Census and Related Population Information, Measure Evaluation, [] (last visited Jan. 30, 2019) (describing the use of repre­sentative sample surveys). The “Census Coverage Measurement,” a postenumeration survey that measures household information, oper­ates as a partial failsafe and can also help convey a more accurate popula­tion count, but broader use of sampling is still necessary to remedy the undercount. 220 Post-Enumeration Surveys, Coverage Measurement, U.S. Census Bureau, https:// [] (last visited Jan. 30, 2019).

D. Engage Private Sector Support to Enhance Accurate Census Enumeration

Both the private and public sectors rely on Census data to operate, plan for the future, and assess the current market. Investment in state-level infrastructure to support the Census can increase engagement. In particular, the private sector can play a key role in funding public education and improving equitable digital access to the Census forms. Indeed, the role of the private sector in supporting Census participation cannot be overestimated. Leading up to the 2010 Census, the philanthropic community used its funding and convening powers to engage a coalition of groups in public-education and Census-participation initiatives. The Funders’ Committee for Civic Participation, for example, gave approximately $40 million in funding for Census outreach, which spawned an extensive grantmaking initiative. 221 Bass & Schless-Meier, supra note 68. In addition, a coalition of civil and human rights groups worked with the Leadership Conference Education Fund and its national partners in support of the campaign to boost Census participation in undercounted communities in over a dozen jurisdictions across the country. 222 See id. For example, “[i]n California alone, a group of funders joined together to distribute around $10 million, five times what the state contributed.” 223 Id.

To ensure that the benefits of technology in the administration of the Census are distributed equitably, local, state, and national service providers might optimize the use of such technology by providing free smartphone applications and subsidizing data usage for online comple­tion of Census forms and accessing information about the Census. In the same way that large private corporations such as Uber and Lyft subsi­dized transportation in certain communities to increase access to the polls during the 2018 midterm elections, 224 In preparation for the November 6th midterm elections, Uber partnered with #VoteTogether and Democracy Works to distribute promotional codes and enable over 15,000 potential voters to receive transportation to and from polls for early voting. See Danielle Burr, Update on Uber Drives the Vote, Uber Newsroom (Oct. 24, 2018), []. In addition, Uber offered a $10 discount on any single ride to the polls on Election Day. Id. Similarly, the ride-sharing application Lyft offered 50% off on all rides to polling stations on Election Day as well as free rides to polls for underserved communi­ties. See The Ride to Vote: Use Lyft to Exercise Your Rights, Lyft Blog (Aug. 23, 2018), []. broadband service providers could provide free data usage to access Census forms at and public education materials about the Census.

In addition, public resources like publicly funded state and local libraries, which stand to benefit from more equitable funding allocation based on an accurate Census count, can help fill the breach by providing digital access to Census forms and assisting community members in fill­ing them out. In fact, there are numerous natural partnerships that state and local governments can leverage to ensure a more accurate Census count if they strategize and think creatively.


As “America’s largest civic event,” 225 Berman, supra note 47. the Census carries the promise of unprecedented inclusion and equal representation on the one hand and continued exclusion and erasure on the other. The path on which we choose to proceed could eventually lead to our triumph as a high-functioning multiracial, multi-ethnic democracy, or perhaps, to our top­pling as a nation. Counting the change that the emerging majority of people of color is producing in our country may not be an exact science, but it can be vastly improved to better inform and prepare for the new demographic landscape of our democracy. Moreover, in light of the stag­nating population growth among white residents, the browning of America should be embraced as a lifeline to its continued vitality. Repeat­ing a history of exclusion and inequity through mutually reinforc­ing discriminatory policies and fraught counting methods will produce dura­ble and corrosive effects. It will also distance us further from the vision of representational equality in which all residents are to be counted—and served—as constituents. We owe it to our democracy to ensure that we do not discount the diversity that makes this country the greatest and most promising democratic experiment to date.