Maia Model (Concerned Citizens for Democracy) - 1st Place (tie), PA House

Judges' Statement

Amid a remarkable and highly competitive set of entries, Maia’s effort rose to the top of the pile as as a meticulously executed and explained experiment in one particular approach to drawing a sound election map.  She explained her intense process for drawing the map in very clear detail in her personal statement.

Personal Statement

The objective of this map was to test the effectiveness of a neutral redistricting methodology that restrains partisan gerrymandering and aims to create a fair map. The method focuses on minimizing splits in political subdivisions and keeping districts within +/-2 percent of the ideal population on a 203-seat House map. The methodology was created by the Pennsylvania non-partisan organization Concerned Citizens for Democracy (CCFD) as a set of neutral criteria for redistricting in order to end gerrymandering.

This methodology creates districts by combining whole municipalities in compact groups to reach the target population. To equalize the population between districts and reduce the deviation, whole municipalities are added or subtracted along the border between districts in a layer by layer manner until the target is achieved. In theory, this method keep districts compact as population groups are moved from one district to another. It should be noted that this methodology favors municipalities that are already compact: non-contiguous and oddly-shaped municipalities are more likely to be split if the population required to connect or keep them compact cannot be balanced.

Once population equality is achieved, districts can then be modified along the borders in the same manner to achieve other goals such as competitiveness or majority-minority districts. (It should be noted that keeping municipalities whole should inherently keep most communities of interest whole). The map does not prioritize competitiveness and was drafted without using partisan data in order to preserve the neutral goal of the method and act as a control for future experiments with this methodology.

Within larger cities, one can see the layer by layer process more easily. Whole wards were treated as townships and combined and split to equalize population. Especially in Philadelphia, this resulted in compact districts and inherently created majority-minority districts. While Pittsburgh’s wards created compact districts, the minority populations are not concentrated enough for majority-minority districts to form by default. This is a case where modifications along district borders could be made to achieve a secondary goal.

While maintaining a 2 percent deviation to prevent extreme “packing”, the objective was for only one municipality to be split along each border and that municipalities be split only once and no more than twice. However, some districts prioritized keeping municipalities intact over the 2 percent deviation - for example, keeping a group of counties together instead of creating additional splits as with Washington, Greene, and Fayette Counties.

Because of the number of districts and varying population densities, it was overall far more practical to keep townships intact than it was counties, even though many counties have a population below the target. There are several counties that are split into more districts than required by their population, and several districts that draw from more than two counties. While this was necessary to meet the target population deviation, one can see on the map that these districts take from neighboring counties in compact layers, minimizing the intrusion (one example: moving south from Erie through Crawford, Mercer, Lawrence, and Beaver Counties).

Overall, the map has 21 majority-minority districts, 36 competitive districts, and an 8.46 percent proportionality deviation. The number of minority-majority districts that formed by default is enough to demonstrate that keeping municipalities whole will inherently protect many communities of interest. Though the map is not very competitive and the proportionality could be better, the map is meant to be a control, a starting place that can be modified to achieve these secondary goals without extreme gerrymandering. Keeping this in mind, the methodology is considered effective in achieving its goals. The DRA app rates the map 71 for compactness and 88 for splitting; 15/203 districts are overpopulated by more than 2%; 16/203 districts are underpopulated by more than 2 percent. Considering that 88 percent of districts meet the target deviation, and the under and over populations are relatively close, the conclusion is that balancing a 2 percent deviation with keeping municipalities whole is practical.