By Ben Williams
As is well-documented by now, states will be receiving redistricting data (aka the P.L. 94-171 redistricting file) from the Census Bureau in August. The current target date is Aug. 16.
Nearly five months behind schedule, the delays at the bureau have had significant impacts on states’ abilities to comply with their own redistricting deadlines. While these delays may provide states with new and unique opportunities to gather public input or make policy changes, the primary challenge states face is addressing the legal catch-22 many states find themselves in.
Responses to the delays have generally fallen into four “buckets”:
- Asking the state’s supreme court to move redistricting deadlines.
- Changing the legal deadline (if still possible).
- Moving the state’s primary dates.
- Using a data source beyond the P.L. 94-171 to redraw their lines.
That last option leads to still further questions. Which alternative sources of data are usable in redistricting? How accurate are they compared to P.L. data? And how can states use alternative data effectively?
These questions and more were answered during the most recent office hours hosted by NCSL’s Elections and Redistricting Program. Kim Brace of Election Data Services walked through the benefits and drawbacks of two options: (a) the five-year estimates produced by the Census Bureau as part of its American Community Survey (ACS) program; and (b) commercially available datasets.
Both the ACS data and commercially available datasets are not enumerations of the population. Instead, they are surveys based on sampling. This and other methodological differences can introduce additional uncertainty into alternative datasets that do not exist in the P.L. files. Also, the P.L. file is designed specifically for redistricting; neither the ACS nor commercially available datasets are designed to apportion political power.
Using alternative data “is not a permanent solution. I look at it as temporary until the P.L. files are released,” said Brace. Once the P.L. files are released, “you’re going to want to swap [the alternative data] out” with the P.L., he elaborated.
And even if a state decides to use alternative data temporarily, they themselves can differ greatly in the numbers they report. While the P.L. data and commercially available datasets are based on population numbers from 2020, the most recent five-year ACS estimates are derived from data gathered from 2015 – 2019, with a midpoint of 2017. Is that recent enough? For areas with rapid growth or depopulation, maybe not.
Each state will decide for itself whether it will wait for P.L. data in August, or go ahead with something else now, understanding the potential pitfalls of alternative data. If you decide to use P.L. data, you can watch Brace’s presentation in full here.
Ben Williams is a policy specialist in NCSL's Elections and Redistricting Program.