If you enjoy drawing district maps or want to give it a try, you’re in for a treat. Daily Kos Election has made our newest data set, which we’ve used to analyze North Carolina’s legislative districts, available to the public. In this post, we'll detail the data we’ve made available in this folder and how you can use it to analyze maps you or others have drawn, or to map election results.
We compiled this data to assess the new maps that Tarheel State Republicans recently drew to replace their illegal gerrymanders, which were struck down by a state court for violating the rights of voters. Our conclusion: They’re still gerrymanders. That’s why we’ve also proposed a much fairer set of nonpartisan maps, which we submitted to the court in an amicus brief.
You can can use our data to dig into the GOP’s plans, our alternatives, or your own, with a very helpful piece of free software called Dave’s Redistricting App. DRA uses the voting tabulation districts (VTDs) that existed in 2010 as its building blocks for drawing districts in North Carolina, but voting precinct boundaries change over time as populations grow or shrink, making it difficult to calculate election results by district when they split precincts.
However, using block-level files from the state's redistricting archives, we have recalculated the results of every statewide partisan election from 2008 to 2018 (plus some from 2004) by the 2010 census blocks and VTDs, which you can find in this folder. You can therefore use this data to calculate the results of any of these elections for any North Carolina maps you encounter.
We’ve also included 2010 census statistics on racial demographics, 2013-2017 census estimates of educational attainment rates among adults and racial demographics among adult citizens (i.e., eligible voters), 2016 primary turnout data by race and party. Along with election results, these racial stats can help you assess whether districts you've drawn would allow black voters to elect their preferred candidates in a primary or general election to comply with the Voting Rights Act.
And for those of you who use GIS software to create your own maps of election results, we have also included the 2010 VTD shapefiles, 2016 and 2018 precinct shapefiles, and the 2016 and 2018 precinct results with absentee, early, and provisional votes assigned to geographic precincts.
To use our data with DRA, your first step is to save the VTDs of any districts you’ve drawn as a .csv file, which you can do in the 2020 version of DRA by clicking the arrow icon pointing upward and to the right on the top banner (under the old version of DRA, it's under the "File" menu on the left side). Then, copy and paste the district column from that .csv into our 2004-2018 VTD-level results file (if you're using the old version of DRA you’ll need to resort the VTD results file on Column C before pasting the districts in). If you've drawn a map at the census block level in GIS software, simply match up your GEOIDs in your block equivalency file with our data sheet, which has the same headers as the VTD sheet.
From there, just sort the whole VTD results sheet or census block results sheet by district and use Excel's subtotal feature under the data tab to tally up the results of each race by district.
Once you have those district-level demographics and results subtotaled, copy them from Excel into a text editor like the Notepad program, then copy them again from Notepad into our district summary display templates for Congress, the state Senate, and the state House depending on which type of map you've drawn. Those templates will auto-format each district's statistics.
If you subtotaled all of the demographic and electoral data, then paste it into the "summary" tab for your type of map starting at cell "AI4." Those tabs have a summary chart of the racial demographics, 2012 and 2016 president, and an average of all 2012-2018 statewide elections below the presidential level, which weights each year equally. The second "detail" tab should auto populate a chart that lists the results by district for every individual election in the data set.
You can see an example of this summary data for our hypothetical nonpartisan North Carolina congressional map below.
Click to enlarge
For each district and the state overall, the first table shows the deviation from the ideal population, the racial demographics of the voting age population in the 2010 census, the 2016 and 2012 presidential results, and a 2012-2018 average of non-presidential results weighted by year. As you can see from the stats, five districts lean Democratic and eight lean Republican. However, the downballot average indicates that three of those red districts (Districts 2, 7, and 11) were more favorable to state Democrats than Democrats running for president and could have been competitive in years like 2012 or 2018 had this map been used instead of the GOP's gerrymanders.
The second table shows the 2013-2017 estimates of racial demographics for adult citizens and educational attainment, 2016 primary turnout by party and race, and Democratic turnout by race. Looking at these two charts in tandem, we can see that the 1st District under this map votes consistently Democratic and has a large black majority among its Democratic primary voters, meaning black voters likely have the ability to elect their preferred candidates in both the primary and general election there.
Click to enlarge