Automated Redistricting Simulation Using Markov Chain Monte Carlo

Fifield, Benjamin; Higgins, Michael J.; Imai, Kosuke; Tarr, Alexander

doi:10.1080/10618600.2020.1739532

Cited by 84 publications

(104 citation statements)

References 34 publications

Supporting

Mentioning

104

Contrasting

Order By: Relevance

“…The units of the variation of information are bits/county. This measure was previously used in other redistricting papers [FHIT18]. 3 The final measure we consider is variation of information divided by the average population change (abbreviated VI/APC).…”

Section: Stability Of Clusterings Over Timementioning

confidence: 99%

Optimal Legislative County Clustering in North Carolina

Carter

Hunter

Teague

et al. 2020

Statistics and Public Policy

View full text Add to dashboard Cite

North Carolina's constitution requires that state legislative districts should not split counties. However, counties must be split to comply with the "one person, one vote" mandate of the U.S. Supreme Court. Given that counties must be split, the North Carolina legislature and courts [Lak02] have provided guidelines that seek to reduce counties split across districts while also complying with the "one person, one vote" criteria. Under these guidelines, the counties are separated into clusters; each cluster holds a number of districts based on its population. Districts may not span clusters, meaning that each cluster forms an independent set of districts in the sense that it can be subdivided into districts without affecting other clusters. In many county clusters, there are more than one district.The guidelines for clustering counties were clarified by the courts in 2015 [Dic15]. In 2017 the districting plans drawn in 2011 for the North Carolina House and Senate were found to be racially gerrymandered. The remedy accounted for the courts' 2015 clustering clarification and redrew some of the clusters and districts for use in the 2018 elections. The enacted set of clusters in both the state House and state Senate were reported to be optimal in that the remedy produced the largest number of county clusters possible while following the outlined procedure. However, no transparent validation of this claim was provided in the public domain.The primary goal of this work is to develop, present, and publicly release an algorithm to optimally cluster counties according to the guidelines set by the court in 2015. We use this tool to investigate the optimality and uniqueness of the enacted clusters under the 2017 redistricting process. We verify that the enacted clusters are optimal, but find other optimal choices. We emphasize that the tool we provide lists all possible optimal county clusterings.We also explore the stability of clustering under changing statewide populations and project what the county clusters may look like in the next redistricting cycle beginning in 2020/2021. In studying the stability of these clusters, we find that their structure may be highly susceptible to small fluctuations in the population; on average approximately one third of the clusters change each year. In addition, we compare the existing guidelines with an alternative interpretation of how counties might be minimally split. As part of this report, we provide code, 1 along with documentation and examples, which may be used by the public to independently verify the North Carolina legislative district clusters during the next redistricting cycle.The report is organized as follows. In Section 1, we lay out the general redistricting problem for North Carolina in light of various laws and legal precedents. We then explain how this leads to the "county clustering" problem and the court-sanctioned procedure for resolving it. In Section 2, we describe our algorithm and give the needed theoretical justification to ensure it produces all of the optim...

show abstract

Section: Stability Of Clusterings Over Timementioning

confidence: 99%

Optimal Legislative County Clustering in North Carolina

Carter

Hunter

Teague

et al. 2020

Statistics and Public Policy

View full text Add to dashboard Cite

show abstract

“…there often exists at least some suspicion that a state's districts have been gerrymandered in some way, even in states with bipartisan or nonpartisan commissions or in states where mapmakers proclaim to be neutral. since it is difficult to estimate the nongerrymandered counterfactual using comparable elections, researchers have recently turned to simulating it using computer-automated districting algorithms (Chen and Cottrell 2016;Cirincione, darling, and O'rourke 2000;Fifield et al 2015;Fryer Jr. and Holden 2011;Krasno et al 2016;Magleby and Mosesson 2018). 4 these algorithms are designed to reproduce the districting process by aggregating Census blocks into a predetermined number of contiguous and equally populated geographic jurisdictions.…”

Section: Measuring the Effect Of Gerrymandering By Establishing The Nmentioning

confidence: 99%

“…For these reasons, studies have generally focused their gerrymandering analysis on just a handful of states. Cirincione, darling, and O'rourke (2000), for example, look only at redistricting in south Carolina; Chen (2017) and Krasno et al (2016) focus on the state legislature in Wisconsin; Fryer Jr. and Holden (2011) limit their analysis to congressional districts in only four states, and Fifield et al (2015) limit their analysis to districts in only two states. Moreover, up until recently, geographically precise data on partisanship spanning all states had not yet been compiled.…”

Section: Measuring the Effect Of Gerrymandering By Establishing The Nmentioning

confidence: 99%

Using Computer Simulations to Measure the Effect of Gerrymandering on Electoral Competition in the U.S. Congress

Cottrell¹

2019

Legislative Studies Qtrly

View full text Add to dashboard Cite

Recent research has leveraged computer simulations to identify the effect of gerrymandering on partisan bias in U.S. legislatures. As a result of this method, researchers are able to distinguish between the intentional partisan bias caused by gerrymandering and the natural partisan bias that stems from the geographic sorting of partisan voters. However, this research has yet to explore the effect of gerrymandering on other biases like reduced electoral competition and incumbency protection. Using a computer algorithm to design a set of districts without political intent, I measure the extent to which the current districts have been gerrymandered to produce safer seats in Congress. I find that gerrymandering only has a minor effect on the average district, but does produce a number of safe seats for both Democrats and Republicans. Moreover, these safe seats tend to be located in states where a single party controls the districting process.

show abstract

“…For instance, an idea for sampling legal districts is to begin with some partition into connected regions (e.g., the challenged map), and then use an MCMC model to explore the space of legal districts. This approach is taken in several papers (Bangia et al 2017;Herschlag, Ravier, and Mattingly 2017;Mattingly and Vaughn 2014;Fifield et al 2017).…”

Section: Introductionmentioning

confidence: 99%

“…Instead, here, "legal maps" refers only to maps that must satisfy some set of constraints. 4 This dataset is available in the redist CRAN package(Fifield et al 2017). 5 This dataset is much smaller than an actual redistricting instance, but it exhibits the phenomenon we wish to highlight that also occurs for the much larger actual redistricting problems.…”

mentioning

confidence: 99%

Understanding Significance Tests From a Non-Mixing Markov Chain for Partisan Gerrymandering Claims

Cho

Rubinstein-Salzedo²

2019

Statistics and Public Policy

View full text Add to dashboard Cite

Recently, Chikina, Frieze, and Pegden proposed a way to assess significance in a Markov chain without requiring that Markov chain to mix. They presented their theorem as a rigorous test for partisan gerrymandering. We clarify that their ε-outlier test is distinct from a traditional global outlier test and does not indicate, as they imply, that a particular electoral map is associated with an extreme level of "partisan unfairness. " In fact, a map could simultaneously be an ε-outlier and have a typical partisan fairness value. That is, their test identifies local outliers but has no power for assessing whether that local outlier is a global outlier. How their specific definition of local outlier is related to a legal gerrymandering claim is unclear given Supreme Court precedent.

show abstract

Automated Redistricting Simulation Using Markov Chain Monte Carlo

Cited by 84 publications

References 34 publications

Optimal Legislative County Clustering in North Carolina

Optimal Legislative County Clustering in North Carolina

Using Computer Simulations to Measure the Effect of Gerrymandering on Electoral Competition in the U.S. Congress

Understanding Significance Tests From a Non-Mixing Markov Chain for Partisan Gerrymandering Claims

Contact Info

Product

Resources

About