Too Large, Too Small, or Just Right? Assessing the Growth of Voter Registration Rates Since the NVRA
48 Pages Posted: 7 Sep 2019 Last revised: 31 Oct 2019
Date Written: September 2, 2019
The number of people registered to vote in the United States since the passage of the NVRA in 1993 has risen by 58% in a period in which the voting-age population has risen 28%. As of 2018, only eight percent of the eligible electorate appears unregistered. What accounts for this dramatic rise in the registration rate and the apparent near-disappearance of unregistered voters? As an initial foray at this question, I pursue two methods to help us understand whether official registration statistics are “too large” and by how much efforts to remove deadwood are successful in removing voters who have moved or died. One method is a simple comparison of official voter registration reports with self-reports to the Current Population Survey. The other method compares list maintenance statistics with demographic patterns that should be driving those statistics. Using the first method, it appears that official statistics over-report actual registration rates by about five percentage points. Using the second method, it is clear that shortcomings in removing deadwood from the rolls comes primarily from removing people who move out of jurisdiction, not dead voters. The conclusion touches on the role that the analysis in this paper can play to contribute to the public discourse about voter registration patterns, and the continuing data challenges in gaining a generalized knowledge about those patterns.
Suggested Citation: Suggested Citation