Knowledge

Training

Can my group receive a training?

While we cannot guarantee availability, we may be able to provide a private training to sufficiently large, non-partisan groups depending on the topic of interest. Please reach out to us at info@redistrictingdatahub.org and send a brief description of your group and the topic you are interested in learning more about.

What trainings are scheduled / coming up?

The best way to be notified about future trainings is to sign up for our mailing list.

Do you have recordings of past trainings?

On the training page, under the heading “Past Training Sessions”. Our trainings are also available on the Redistricting Data Hub YouTube channel

Process

What is the difference between reapportionment and redistricting?

Reapportionment is the act of reallocating congressional seats across states, while redistricting is the act of drawing the boundaries of these districts. They are inter-related but distinct processes. Sometimes, however, “apportionment” is used to mean “redistricting,” such as Arkansas’s Board of Apportionment, which is charged with legislative redistricting.

Who is responsible for redistricting in my state?

Most commonly, committees in the state legislature are responsible for legislative and congressional redistricting, and governors must approve.

A number of states have also adopted commissions, although these come in many forms, ranging from independent, citizen commissions, such as those found in California and Michigan, to small, politician commissions, such as those found in Arkansas and New Jersey. Some states have advisory commissions, who can propose but not pass maps.

Since each state ultimately decides who will be responsible for redistricting, there are many variations from these general rules. And it is important to remember that different bodies might be responsible for legislative and congressional redistricting. You can find out who is responsible for redistricting in your state from All About Redistricting or the Brennan Center.

Where can I find the most recently enacted map in my state?

Our team tracked down all enacted state legislative and congressional maps from the 2021 redistricting cycle; we continue to monitor ongoing litigation and rehost maps as they are adopted. To download the most recent maps and any other maps in place back to 2011, and set Filter by Type of Data to “legislative boundaries.” If you are looking for a very recently enacted map that is not yet available on our website, reach out to the Help Desk. We may be able to direct you to an official source for the data, or a reliable copy of the map in Dave’s Redistricting App.

How do I decide what mapping tool to use?

Given the variety of needs for those using mapping tools, we do not recommend a particular tool above all others. Instead, the RDH has a Choose Your Own Mapping Tool that breaks down the various mapping tools and their features, including available languages, cost, support, geographic coverage, and much more.

Criteria

What criteria does my state require for redistricting?

All legislative and congressional redistricting plans must comply with two federal requirements: equal population and the Voting Rights Act of (1965). In addition, states may adopt their own criteria. The most common criteria are compactness and contiguity, followed by respecting communities of interest, political, and geographic boundaries. Visit your state page on the All About Redistricting website to see what criteria are in effect for your state.

What is compactness? How is it measured?

Compactness is a common state-imposed requirement for redistricting plans. The language varies from state to state, but generally requires redistricting plans to contain geographically compact districts. There are several measurements of compactness, including Reock, Polsby-Popper, length-width, and others.

What is contiguity?

Contiguity is a common state-imposed requirement for redistricting plans. The language varies from state to state, but generally requires redistricting plans to contain districts that connect geographically with all other parts of itself. There may also be requirements around “point contiguity,” in which one part of a district’s boundaries touch only at a single point, and in district boundaries passing through water.

What is a community of interest?

Respecting communities of interest is a common state-imposed requirement for redistricting plans. The language varies from state to state, but generally requires redistricting plans to consider or respect “communities of interest” when drawing district boundaries. A community of interest is sometimes defined more specifically, to include examples or considerations that may (or may not) be taken into account.

What is the equal population requirement?

In Baker v Carr (1962), the U.S. Supreme Court ruled that redistricting was a justiciable question under the equal protection clause of the Fourteenth Amendment. They subsequently ruled that legislative (Reynolds v. Sims 1964) and congressional (Wesberry v. Sanders 1964) districts with unequal populations were in violation of the equal protection clause.

In the case of congressional districts, the goal is equality as “nearly as practicable,” meaning deviations as small as one person across districts. In the case of legislative districts, the goal is “substantial equality,” which has been defined as deviations of less than 10%. Thus, all states must draw redistricting plans that are (roughly) equal in population, for both congressional and legislative – upper and lower – districts.

Automated Redistricting

How do automated redistricting (AR) algorithms work?

Automated redistricting algorithms can be separated into two broad categories based upon their use case. The first category is automated algorithms that are used to replace the work of human map drawers by creating an optimized map under certain criteria. The most common type of these algorithms, also referred to as optimization algorithms, maintain contiguity and population balance of the districts and try to maximize the “compactness” through some measure of shape. One difficulty in these sorts of algorithms is including multiple criteria (say, equal population, compactness and VRA laws) and explicitly defining the tradeoffs among the assorted criteria.

The second category is automated algorithms that are used to analyze redistricting plans by creating a large numbers of plans, also called an ensembles. These algorithms then compare a particular plan relative to the set of other plans. One difficulty in these sorts of algorithms is that there are a massive number of possible redistricting plans for a particular state or area, which means that an ensemble cannot include every possible redistricting plan. For the ensemble analysis to be fair and meaningful, the ensemble must effectively sample the set of all possible redistricting plans, which although theoretically possible, can be difficult to achieve in practice.

How can automated redistricting ensembles be used to analyze maps / identify gerrymandering?

There are four different ways in which ensembles are used to analyze maps:

• Provide evidence for partisan gerrymandering by running an outlier analysis.
• Study the effects of enacting redistricting reforms by changing the prioritization of different things when constructing an ensemble of maps (ex. making minimizing county splits one’s highest priority)
• Analyze the effects of gerrymandering by creating a “non-gerrymandered” counterfactual and compare what happens in that world to what happened in our world.
• Evaluate proposed maps by analyzing past election results and generating predictions for future election results.

How “realistic” are the maps generated by a particular Automated Redistricting (AR) algorithm?

Given the sheer variety of automated redistricting algorithms and techniques, there is no single answer to this question. In general, however, it is important to evaluate the inputs and methodologies of an AR technique to have a sense for how useful it’s outputs (map) may be. While not exhaustive, some questions to ask include:

• What unit are the maps being built out of?
• What are they using as a starting point?
• How do they construct the district through the problem?
• What metrics does the algorithm use to quantify criteria?
• How does the algorithm prioritize these different criteria?

Gerrymandering

Is racial gerrymandering legal?

Racial Gerrymandering

Racial gerrymandering can be defined as drawing districts in order to minimize the voting power and representation of people belonging to a particular racial, ethnic, or language minority (protected) group.
The Voting Rights Act (VRA) of 1965 prohibits redistricting plans that deny or abridge a protected group’s right to elect representatives of choice.
The U.S. Supreme Court has also held that race cannot be the “predominant” factor in redistricting, unless there is a compelling reason, such as compliance with the VRA (Shaw v. Reno 1993).
The All About Redistricting website provides an excellent summary of how redistricting must (not) take race and ethnicity into account.

What is racial voting bloc or racially polarized voting?

Racial Gerrymandering

Section 2 of the Voting Rights Act (VRA) prohibits minority vote dilution, as demonstrated by totality of the circumstance of the local electoral process. When considering the totality of circumstances, courts often defer to a list of specific considerations known as the “Senate factors.”

One of these factors is “the extent to which voting in the elections of the state or political subdivision is racially polarized.” The analysis for this factor is referred to as racial voting bloc or racially polarized voting (RPV) analysis. RPV analyses require data about election results and the race of voters.

Ideally, we would know the race of voters and how they voted. How people vote in congressional and legislative elections are kept secret, however, and thus unknown. So some statistical effort is required to infer the race of voters and how they voted. There are three statistical methods that are commonly presented as RPV analyses in Section 2 cases: homogenous precinct analysis, ecological regression, and ecological inference.

What software can be used to conduct racially polarized voting (RPV) analyses?

Racial Gerrymandering

There are three packages available in R that conduct racially polarized voting analyses:

1. ei, developed by Gary King and Molly Roberts, which conducts the method of ecological inference developed by King.
2. eiCompare, developed by Matt Baretto, Loren Collingwood, and colleagues, which compares the results of Goodman’s ecological regression, ecological inference, and the “rows by columns (RxC)” method (Rosen et al.). The RxC method allows for analysis of elections involving more than two racial groups or candidates.
3. eiPack, developed by Olivia Lau, Ryan T. Moore, and Michael Kellerman, which provides methods for analyzing RxC tables using extreme case analysis, ecological regression, and Multinomial-Dirichlet ecological inference models.

There is also an online shiny app that can be used to run EI, created by the Metric Geometry and Gerrymandering Group through their Voting Rights and Data Institute.

Is partisan gerrymandering legal?

Partisan Gerrymandering

Partisan gerrymandering can be defined as drawing districts in order to maximize the number of seats a single party will win. The U.S. Supreme Court has said that excessive partisanship in the redistricting process is unconstitutional, but that they are unable to hear such claims because there is no agreed-upon standard for what constitutes excessive partisanship. However, many states have adopted criteria or even constitutional language designed to prevent or mitigate partisan gerrymandering. The All About Redistricting website provides an excellent summary of these rules.

What is prison-based gerrymandering?

Prison Gerrymandering

In 2020, the U.S. Census Bureau counted inmates who were in jail or prison on April 1 as residing at the facility in which they are currently incarcerated. Because inmates are not necessarily confined in a facility close to their home; because the demographics of inmates are not necessarily reflective of the demographics of residents living near the facility; and because the vast majority of states prohibit incarcerated people from voting, the counting of inmates at facilities for redistricting purposes frequently distorts representation.

As a result, several states have prohibited prison-based gerrymandering, by “reallocating” inmates to their last known address. States differ widely in which inmates are reallocated, whether or how inmates without a last known address are reallocated, and whether reallocation is performed for legislative and/or congressional redistricting.

General

Who are you funded by?

The below information is all located on our About Us page. The Redistricting Data Hub is a project of the Fair Representation in Redistricting Initiative, a project of New Venture Fund (NVF). NVF has contracted with HaystaqDNA—a political data consulting firm with a track record of success in supporting nonpartisan redistricting—to provide the publicly-available data and services. There is a firewall between the Redistricting Data Hub work and services and any other work provided by HaystaqDNA to any of its other clients. Funding for the Redistricting Data Hub comes from the Fair Representation in Redistricting Initiative. The Initiative is guided by a Fair Representation in Redistricting Advisory Committee composed of representatives of the following 18 initiative funders:

• Annie E. Casey Foundation
• Bauman Foundation
• Bernard & Anne Spitzer Charitable Trust
• California Community Foundation
• Democracy Fund
• Ford Foundation
• Grove Foundation
• Gund Foundation
• Joyce Foundation
• JPB Foundation
• Mary Reynolds Babcock
• New York Community Trust
• Open Society Foundations
• Resilient Democracy Fund
• Rockefeller Brothers Fund
• State Infrastructure Fund
• Wallace H. Coulter Foundation
• W.K. Kellogg Foundation

Funding for state and local data in 5 southern states is provided by the Southern Poverty Law Center

How do you support journalists covering redistricting?

The RDH maintains a help desk where we respond to any redistricting or redistricting data related inquiries. Some types of inquiries that we’ve handled include questions about state redistricting processes / timelines, more information on certain datasets or questions about where to find a particular piece of data. Please send us an email at info@redistrictingdatahub.org to get in touch with the help desk.

What is in your user agreement?

You can read our Terms and Conditions of Use and our Privacy Policy. The data on our website is for noncommerical use only and that the RDH must be credited alongside any uses of the data.

Does my state require use of the decennial census data for redistricting?

Decennial census data

The National Conference of State Legislatures has put together a guide on the subject; you can see whether your state requires decennial census data for legislative and/or congressional redistricting on their Redistricting and Use of Census Data page.