Welcome to Point/Counterpoint, where Elections Daily writers will provide differing perspectives on various topics. This first edition pits Eric Cunningham and Adam Trencher against each other on the topic of Arizona’s congressional map, which was passed by the Independent Redistricting Commission in 2011.
This map has been a topic of heated debate since it was proposed, with Republicans arguing it is a partisan gerrymander and Democrats arguing it’s a fair map. In this debate, Eric will be positing that the map is a gerrymander while Adam will argue that it is not. You can find Eric Cunningham’s counterargument at this link. Without further ado, let the debate begin!
With the Supreme Court’s ruling in Rucho v. Common Cause finding partisan gerrymandering to be nonjusticiable, the search for fair maps nationwide seems more elusive than ever. However, this ruling hasn’t stopped states, and its voters, from taking matters into their own hands. As a matter of a fact, long before this ruling, that’s exactly what Arizona voters did when they passed Proposition 106 in 2000, which created the Arizona Independent Redistricting Commission (AIRC), and tasked it with crafting congressional and legislative maps that were compact, represented communities of interest, preserved jurisdictions like towns and counties, and created competitive districts.
However, despite being drawn by this independent commission, Arizona’s congressional maps are much maligned as a partisan gerrymander. This begs the question though: Is Arizona’s congressional map actually a gerrymander, or a map that truly reflects that states diverse communities and creates competitive districts?
Proposition 106 establishes both an initial map design, and an ordered list of goals which dictate adjustments to the initial map design, so it is a useful starting point to see if Arizona’s congressional map fulfills those goals. As stated in Proposition 106, “the commencement of the mapping process for both the congressional and legislative districts shall be the creation of district of equal population in a grid-like pattern across the state.” In accordance with this, the committee ordered its mapping consultants to generate grid maps, and adopted the below map as a starting point, which was created by beginning in the southeast corner and continuing clockwise.
Using this grid map as a starting point, it’s only logical to view the actual congressional map for evaluating whether it fulfills the Proposition 106 goals in the context of changes to this basic design. Essentially, are the necessary changes to make this initial map fulfill those goals reflected in the actual map?
VRA and population
The first goal to make sure the districts are compliant with the U.S. Constitution and Voting Rights Act. For the purposes of Arizona congressional redistricting, a Constitutional map must have essentially no population deviation between districts, and a VRA compliant map must have 2 congressional districts in which Hispanics comprise at least 55% of the voting-age population and are thus consistently able to elect a representative of their choice. The initial grid has a maximum deviation of 1 person from the average in each district, and the actual Congressional map managed to eliminate that and have 9 equally populated districts. Thus, the Arizona congressional map clearly complies with the U.S. Constitution.
Onto VRA compliance, while there are no available demographic stats for the initial grid, it can be reasonably inferred from visually inspecting the map that it is not compliant with the VRA. One could infer this because districts 1 and 2 appear to near evenly chop the Hispanic portions of the Tucson area, and the remainder of both districts are sufficiently white that neither would qualify as a VRA district; districts 3, 7, and 8 appear to chop the Hispanics portions of the Phoenix area such that it is questionable whether even one of them qualifies as a VRA district, and very unlikely that two of them would qualify.
By contrast, the Arizona congressional district map has the required 2 VRA districts in districts 3 and 7, which are both over 55% Hispanic in terms of the voting-age population, and thus in compliance with the VRA. This compliance was affirmed by the map obtaining preclearance from the Justice Department. From this, it is manifest that Arizona’s congressional map meets the first goal established by Proposition 106.
The second goal is that the districts have essentially equal population. This is redundant because Wesberry v. Sanders establishes that all congressional districts must have essentially no population deviation to comply with the U.S. Constitution, so it’s already covered by the first goal, and shown to have been met there.
The third goal is that districts be “geographically compact and contiguous.” The Arizona congressional map is particularly criticized in this regard, with the 1st district described by our Editor-in-Chief as “a monstrosity that ignores all guidelines of compactness.” However, while it’s easy to be incredulous at the shape of the 1st district, or at the map in general, while eyeballing it, it’s much harder to show it statistically to be lacking in compactness, because it isn’t.
One of the most commonly used and respected tests of compactness is the Reock score, which is simply the area of a district divided by the area of the smallest circle which fully encloses it. Higher scores are better, as they indicate a more compact district. The average Reock score of the districts in the initial grid is approximately .50, and the average Reock score of the districts in Arizona’s congressional map is approximately .45.
You might have two observations to this, which are, “wait it got less compact from the initial grid,” and, “that doesn’t sound very compact.” Regarding the first observation, it was necessary for the districts to become slightly less compact to comply with the VRA. For example, district 3 on the initial grid had to drop almost all of the Pinal portion and the Maricopa portion outside of the West Valley in exchange for parts of the Tucson area to become a VRA district for the actual map, with a significant drop in compactness as a result. Additionally, parts of the map became more compact. The “monstrous” 1st district is more compact than the 5th district on the initial grid that it was based off.
Regarding the second observation, the Arizona congressional map is actually one of the most compact in the nation based on the average Reock score of its districts. Had the initial grids been adopted, it would have been the most compact map in the nation. The actual congressional map isn’t far behind, being edged out only by Nevada and Nebraska, both of which have far fewer people and distinct communities of interest to account for. In fact, had the “monstrous” 1st congressional district been its own state, it would have been tied with Nevada for the most compact map.
While it’s easy to be incredulous at aspects of the map without statistics, the statistics show that Arizona congressional map’s compactness is “monstrous” only in the same way that Pete Alonso’s ability to hit home runs is “monstrous,” and that the map complies with the third goal of Proposition 106.
Communities of interest
The fourth goal of Proposition 106 is the preservation of communities of interest. The most critical question here is: “what is a community of interest?” Communities of interest are notoriously hard to define, but Strategic Telemetry, a mapping consultant to the AIRC, does make an important point on defining them in their response to the state’s request for proposal, which reads, “there is no single definition of a community of interest. It is a concept best defined by those who live in the affected areas.” Thus, when considering if the map preserves communities of interest, it’s best to look at public feedback to draft maps and maps submitted by outside parties.
Looking through public feedback, the most relevant comments seem to be that rural areas generally want their counties preserved, the Oro Valley is a community of interest, Ahwatukee is a community of interest, Flagstaff-ers do not want to be put with Prescott, and Prescott wants the Verde Valley to be put with Flagstaff. These comments are mutually exclusive, as residents of Yavapai County want the county preserved in one district, but also want Verde Valley, which is part of the county, put with Flagstaff in Coconino County, yet residents of Flagstaff do not want to be put with Prescott, in Yavapai County. To fulfill any two of those three goals would preclude fulfilling the third.
Now, looking at outside organization maps, Pinal County wanted the bulk of the county to be put with the Eastern part of the state, and the Navajo Nation wanted to be drawn into the most Native American district possible. The initial grids preserved Mohave and Yavapai counties, but failed to preserve Pinal County, split the Oro Valley and Ahwatukee, put Flagstaff with Prescott, put the Verde Valley with Flagstaff, put the bulk of Pinal County with the Tucson area, and splits heavily Native American areas into 4 different districts.
The AIRC resolved most of these issues in accordance with public feedback and outside organization maps. The Oro Valley and Ahwatukee are preserved in the 1st and 9th congressional districts, respectively. Flagstaff and Prescott are put in separate districts, and the Verde Valley is put with Flagstaff. The bulk of Pinal is combined with Eastern Arizona. Most of the heavily Native American areas are combined into the 1st district, with the bulk of the remaining areas ending up in the 3rd district. However, Mohave, Pinal, and Yavapai counties are all split, though Mohave County is split to keep heavily Native American areas together, Pinal is split in one place to keep a certain tribe together, and in another place to create population equality, and Yavapai County is split to put the Verde Valley but not Prescott with Flagstaff.
Ultimately, it is clear that the Arizona congressional map respects the communities of interest that were presented to the AIRC, and thus fulfills the fourth goal of Proposition 106.
The fifth goal is that maps utilize “visible geographic features, city, town, and county boundaries, and undivided census tracts.” Visible geographic features are hard to quantify, and census tracts doesn’t mean a thing to the average Arizonan, so let’s focus on municipal and county lines. Initial grids don’t have information on municipal splits but can be expected to have far more since municipalities aren’t taken into account.
The Arizona congressional map splits 19 municipalities out of 464 in the state. However, many of these municipalities are discontinuous, and only 12 are split such that at least 0.01% of the population is in a district other than the majority of the municipality’s residents. These splits are reasonable, and generally done to comply with the VRA, like splitting Tucson and Glendale, or to create population equality.
As to splitting counties, 7 of 15 counties are split, but all of them can be explained by achieving more preeminent goals. Maricopa and Pima Counties have more people than any one district could hold and must be split to comply with the Constitution. Yuma is split to comply with the VRA, and while the district could likely comply without splitting the county, it likely only could at the expense of splitting something else. Mohave, Pinal, and Yavapai Counties are split for the reasons described above. Gila County is split as to put the portions within Native American reservations with the 1st congressional district, but the remainder of the county with the 4th.
However, since Pinal County is also split between the 1st and 4th districts, this begs the question of why not keeping Gila whole by putting all of it in the 1st district and moving some of Pinal County in the 4th. There is no clear answer to this, but the most likely one is that this would necessitate splitting a Pinal municipality, and the AIRC decided that keeping the municipality whole was more important than keeping Gila County whole. Regardless, its clear that the Arizona Independent Redistricting Commission respected municipal lines, and had a less than perfect result with county lines, but only in the service of more preeminent goals, so it fulfilled the 5th goal of Proposition 106.
Finally, the 6th goal is to create competitive districts. For this, “create competitive districts” should be interpreted as creating districts that data from prior to 2011 would suggest is competitive, because the AIRC is not known to contain any time-travelers or mind-readers. The initial grids have no available partisan data. As for the actual map, a key question is: “Does competitive mean a toss-up when the state, or the nation, has a near tied result?” Proposition 106 isn’t specific on this, so it’s worth considering both cases.
For “competitive when the nation is tied,” there isn’t a lot to work with because 2008 results are influenced by John McCain’s spot on the ballot, and 2004 results aren’t universally available. However, looking at 2008 results, 7 of the 9 districts voted to the right of the nation, and McCain won 6 of them. The closest districts were 4 (9th), 8 (2nd), and 10 (1st) points to the right of the nation. This would indicate that in a tied national election, all of those districts would go for the Republicans by quite a bit.
As for “competitive when the state is tied,” the AIRC used a number of results and indices for measuring partisanship, and the most useful one being Competitive Index 2, which is the average of the average statewide election result in 2008 and the average statewide election result in 2010. The statewide number on CI2 was Republican by 8.4%, and those three above districts were Dems by 2.2 (9th), tied (1st), and GOP by 1.2 (2nd). This would suggest that in a tied statewide election, all of those districts would go for the Democrats by a quite a bit.
These two conflicting definitions and results would suggest that either party would have something to gripe about depending on which definition they use, which is certainly an endorsement of its competitiveness. Thus, one could reasonably conclude that Arizona’s congressional districts are competitive, and thus fulfill the 6th goal of Proposition 106.
Arizona’s map meets Proposition 106’s goals. Arizona’s map complies with the Constitution and the VRA, is compact, preserves communities of interest, municipalities, and counties where possible, and creates competitive districts. This criteria from Proposition 106 is sufficient such that a gerrymandered map would be unable to meet it, so Arizona’s map is not a gerrymander. As a matter of a fact, considering its success in so many regards, the Arizona map is a bright spot for congressional mapmaking, and should serve as a model to other states both in terms of process and outcome.