Keywords

1 Introduction

Drawing election districts and plans to comply with the law confronts one with many technical issues in compiling, interpreting, and presenting Census Bureau and other data. Data define, identify, and count separate minority populations protected under the law; identify and count those members of each population who qualify as eligible to vote; and show their spatial concentration within a municipality or state. Chapters 8, 9, 10, 11, 12 and 13 highlight numerous issues that arise in such contexts and illustrate ways to address them.

2 Chapter Overviews

Chapter 8 Public Involvement in Balancing Traditional Districting Criteria has particular relevance for independent citizen redistricting commissions and citizen participants in the districting process. It highlights three common concerns: avoiding minority vote dilution, preserving communities of interest, and drawing reasonably compact lines. The case study recounts the public process through which the city of Waterbury, CT agreed upon and enacted a new five-district city aldermanic districting plan in 2015. Successive public commission meetings over several months accommodated a lengthy process of negotiation among citizen groups with different agendas. The outcome was a plan that addressed these three concerns.

Chapter 9 Characterizing Minority Voting Strength in Locally Diverse Contexts illustrates various ways that local demographic contexts shape minority voting strength. Its five case studies exemplify two contextual dimensions: spatial and compositional. The spatial context denotes a population’s distinctive geographic pattern of residence within a community (e.g., concentrated vs. scattered). The compositional context is the population’s distinctive demographic makeup (e.g., citizenship, age structure). The spatial arrangement and demographic composition of a community’s racial and ethnic minority residents may impose practical limitations on drawing districts. Each case study exemplifies how those limitations can impinge upon the possibility of meeting the first “Gingles” precondition (geographic compactness). Orange County, FL features the spatial complexity of Hispanic residential patterns there, reflecting Hispanics’ diverse national origins and derivative political persuasions which may bear on Hispanics’ overall political cohesiveness . Santa Monica, CA features the incongruent residential patterns of that city’s Hispanic and Black residents, which limits the possibility of forming even a crossover or influence district, let alone a minority-coalition district (see “5.2 Types of Districts” above). SeaTac, WA illustrates the practical limitations that derive from this city’s residential patterns . Blacks, Hispanics, and Asians are present in significant numbers but insufficiently concentrated in any one area to make a majority in any potential single-member control district. In all three places, superficial demographic appearances mask underlying political and practical realities.

Two further cases illustrate how a compositional effect (noncitizenship) may exaggerate a group’s apparent presence among eligible voters despite its substantial demographic presence among all residents of a community. Gainesville, GA features a fast-growing Hispanic population that comprised 42% of the city’s residents but only 12% of its eligible voters. Here, we illustrate a simple, elegant method for ruling out any prospect of satisfying the first Gingles precondition as mathematically impossible. This method can circumvent the lengthier trial-and-error exercise necessary to rule out (or confirm) that possibility. The Coppell Independent School District in Texas features a population where Asian Americans comprise 24% of all residents but only 18% of eligible voters, many of them concentrated in several noncontiguous areas within the school district. Here, both spatial and compositional factors preclude satisfying the first Gingles precondition.

Chapter 10 Unmasking “Packing” and “Cracking” for Racial or Partisan Purposes illustrates approaches to documenting two forms of vote dilution --“packing” and “cracking”--whether racial or for a partisan purpose. Variants of these approaches may be applicable to forthcoming instances where post-2020 Census redistricting poses issues of partisan or racial gerrymandering. Dallas County has a five-member governing body, four of whom are each elected by district. This case study illustrates the steps in detecting racial vote dilution in redrawn districts enacted after the 2010 census. We show how the reconfiguration of district boundaries effectively disenfranchised at least one of every nine White voters and one of every 10 Black voters countywide. Maryland’s 6th Congressional District, reconfigured as part of the State’s 2011 redistricting plan, is a case study in gerrymandering for a partisan purpose. Here, “packing” and “cracking” flipped the district from Republican to Democratic control. This case study illustrates the steps in documenting: (1) systematic dismemberment of an existing district through the excessive interchange of territory and population; (2) disregard of existing communities of interest; and (3) a partisan aim–here, replacing Republican registered voters with Democrats. It also unveils a worrisome legacy: further residential separation of new outsiders from the long-established populace, potentially undermining commonalities of interest tied to place.

Chapter 11 Integrating Administrative, Political, and Statistical Geography illustrates several specialized uses of census data in conjunction with political and administrative data. In each instance, the problem addressed was aligning different geographies. A county or city defined by census geography (e.g., census tracts and blocks) also may be organized for administrative purposes by school attendance areas, voting precincts, transportation zones, neighborhoods, and so forth. Aligning standard census geography with these other geographies is tricky and may necessitate some form of approximation. This problem arises in drawing district boundaries that will align with existing administrative boundaries (such as neighborhood or election precinct boundaries); in analyzing prior elections within a community to determine the degree of voters’ cohesiveness or the presence of white block voting; in aligning historical census geography and data from prior decades to address the “totality of circumstances” referenced in Senate Factor 5. The city of Pasco, WA adopted single-member districts to elect its six city council members. In drawing six newly-established election districts, priority was given to maintaining existing voting precincts to simplify the administration of future elections. The boundaries of census blocks aligned with those of existing election precinct boundaries, with a single exception. To resolve this issue, we used Google Maps to visualize and document a straightforward technical adjustment. The city of Santa Monica, CA (as part of a legal defense) needed to reconstruct its demographic past using historical census counts of the city’s minority population. Doing so required matching up contemporary and historical census tract geography and also aligning contemporary definitions of “Hispanic” and “White nonHispanic” with their imperfect historical counterparts from prior censuses. Dallas County features the integrated use of historical 1960 Census gross migration flow data and contemporary Census PUMS (Public Use Microdata Sample) data to profile subpopulations relevant to Senate Factor 5. It outlines a method for gauging the extent to which effects of earlier policies and practices persist (or “linger on”) among members of a contemporary population.

Chapter 12 “False Positive” Majority Minority Election Districts in a Statewide Congressional Plan illustrates a handy method for calculating the probability that one (or more) purported majority minority districts is a “false-positive,” i.e., actually not over 50%. This method can be applied where a multidistrict election plan features several majority minority districts, each calculated to have slender racial/ethnic majorities. A plan boasting several such districts in which the estimated number of Blacks (or Hispanics) barely exceeds some threshold (e.g., 50%) may well be vulnerable to challenge by an opposing party if at least one of the districts is likelier than not “below 50%” (a statistical “false positive”). A step-by-step illustration shows how to gauge that possibility for a given plan, along with practical guidelines for evaluating any plan’s vulnerability to challenge on this important technical point.

Chapter 13 Accounting for Prisoner Populations features a case study of the issues introduced by large prison populations. It explores the prospects for drawing a majority minority election district under hypothetically different rules for counting (or excluding) prisoners as “residents” and as “eligible voters.” Criminal disenfranchisement laws, which strip voting rights from people with past criminal convictions, are in a state of flux (as of this writing, laws differ from state to state). For districting purposes, prison populations are distinctive: they are geographically concentrated populations of voting-age persons, typically skewed toward minorities. Disenfranchisement of felons has been upheld by the Supreme Court. Depending upon the state, it may be that imprisoned populations will count both as residents and as eligible voters; or as residents (but not as eligible voters); or as neither. For Gulf County, FL, which elects its five County Commissioners by district, we illustrate the steps in evaluating the possibility of forming a majority-Black district that would satisfy the first necessary “Gingles” precondition (compactness). The fact that nearly one-fifth of all Gulf County residents are imprisoned convicted felons introduces considerable ambiguity. A majority-Black district could be formed by connecting 3014 imprisoned residents tightly concentrated in one area to several smaller concentrations of Black residents living elsewhere in the County. In Florida, however, imprisoned populations count only as residents, not as eligible voters. Supplementing census data with the prison’s administrative data on current prisoner counts clarifies the issue: Under then-current law, the total population of Black eligible voters in Gulf County (minus those imprisoned) would fall short of the necessary Black majority among non-imprisoned adults in any possible single-member district. A change in the law, however, could alter that conclusion with the mere stroke of a pen.