Is the EAVS a Reliable Guide to Voter List Maintenance?
66 Pages Posted: 5 Sep 2018 Last revised: 12 Oct 2018
Date Written: August 25, 2018
Abstract
This paper examines the reliability of voter registration and list-maintenance data reported in Section A the EAC’s Election Administration and Voting Survey. The EAVS is the most comprehensive nationwide source of such data, and is commonly used in academic studies and legal proceedings. The EAVS project has a long history, tracing its origins to reporting requirements under the National Voter Registration Act in the 1990s. Examining data releases from 2010 to 2016, we show that top-line items tend to be complete, with the greatest amount of missing data pertaining to the removal notification process. Sub-aggregates also tend to be consistent with reported top-line statistics, with notable exceptions. The area where the data are the least consistent is where they can be assessed using a stock-and-flow model of voter registration. Applying data from Section A to a basic stock-and-flow model reveals that the model virtually never balances for any local jurisdiction. That is, the total number of voter registrations at the end of time period t is generally not equal to the number of registrations at the end of period t-1 plus additions and minus removals during the period. The paper concludes by cautioning against taking the Section A data “as-is,” especially in litigation. Challenges to the quality of the data that were first identified in the 1990s persist to this day. The paper’s conclusion also provides some thoughts about improving the quality of Section A data moving forward.
Suggested Citation: Suggested Citation