A Quarter-Million Questions: New York’s Voter Rolls Show Dramatic Spike in Duplicate Registrations in May
by Floyd Patrick, Project CIVICA Data Analyst and Kim Hermance, President, Co-Founder, and Co-Director, Project CIVICA.
May 2026* Election Integrity Score Sheet Reveals Major Increase in Duplicate Voter Records While Other Data Quality Concerns Persist
Every election begins long before a ballot is cast.
It begins with a voter registration database.
Whether someone votes in person, by absentee ballot, or during early voting, the integrity of every election depends upon one simple principle: the voter rolls must be accurate.
Federal and state law recognize that principle. The challenge is maintaining a voter registration system that is both accurate and fair—one that removes registrations that no longer belong on the rolls while protecting every eligible citizen from being improperly removed.
Each month, Project Civica analyzes publicly available New York voter registration data to identify trends, monitor list maintenance activity, and evaluate the overall condition of New York’s statewide voter registration system.
The May 2026 Election Integrity Score Sheet reveals several encouraging improvements—but also raises important questions that deserve public explanation.
Most notably, the number of duplicate voter registrations identified in the statewide database increased by nearly 286,000 records in a single month.
While duplicate registrations do not automatically indicate improper voting, they represent a significant data-quality issue in a statewide voter registration system that federal law requires to maintain unique voter records.
The Law Requires Accurate Voter Rolls
Congress addressed voter registration accuracy through two major federal statutes.
The Help America Vote Act (HAVA) requires every state to maintain a single, uniform, official, centralized computerized statewide voter registration list.
The National Voter Registration Act (NVRA) requires states to conduct reasonable, nondiscriminatory voter list maintenance programs that remove registrations belonging to deceased individuals and voters who have moved while protecting eligible voters from being improperly removed.
New York law likewise establishes procedures for cancelling voter registrations due to death, relocation, duplication, or other legally recognized reasons.
The purpose of these laws is straightforward:
Election officials must maintain voter rolls that are accurate, current, and transparent.
Overall Registration Continues to Grow
New York added nearly 25,000 active voters during May.
Active voters
April: 13,604,914
May: 13,629,898
Increase: 24,984
Inactive registrations declined slightly.
Inactive voters
April: 927,036
May: 925,841
Decrease: 1,195
Meanwhile, the cumulative number of voter records designated as purged continued to grow.
Total Purged Records
April: 10,064,499
May: 10,099,775
Increase: 35,276
During May alone, election officials removed 59,577 registrations, an increase of nearly 16,000 removals compared to April.
Greater transparency regarding the reasons for these removals—whether deaths, relocations, duplicate registrations, or routine maintenance—would help the public better understand how list maintenance is being administered.
Party Registration Trends
Party registration changes remained relatively modest.
Democratic registrations increased by 7,116
Republican registrations increased by 1,907
Voters with no party affiliation increased by 17,294
The largest growth occurred among voters choosing no party affiliation, continuing a trend seen in recent years.
The Largest Finding: Duplicate Registrations
One statistic stands apart from every other number in this month’s report.
New York’s duplicate voter records increased from 170,720 to 456,677.
That represents an increase of 285,957 duplicate records—approximately 167.5 percent in a single month.
This is by far the most significant development in the May data.
Duplicate registrations are not, by themselves, evidence of illegal voting. People may appear as duplicates for a variety of administrative or data-related reasons. Nevertheless, HAVA requires states to maintain unique voter records within a centralized statewide registration system, making a change of this magnitude worthy of public explanation.
Several questions naturally arise:
What caused this dramatic increase?
Was a new matching methodology implemented?
Were additional categories of possible duplicates identified?
Are these confirmed duplicate registrations or potential matches requiring further review?
Without additional information from election officials, the public has no way to determine what accounts for this substantial increase.
Interstate Registration Questions Continue
Cross-state registrations remain another area deserving continued attention.
Project Civica identified:
88,174 apparent New York/New Jersey duplicate registrations
13,169 apparent New York/Florida duplicate registrations
Florida remains one of the most common destinations for relocating New York residents, making interstate voter list maintenance particularly important.
Although interstate duplicate registrations are not inherently improper, they underscore the importance of timely interstate data comparisons and list maintenance efforts.
Duplicate State Identification Numbers Improved
One encouraging development involved duplicate State Board of Elections identification numbers.
April: 1,685
May: 274
This represents a reduction of 1,411 records.
If accurate, election officials should explain what corrective actions resulted in this significant improvement so that similar practices can be applied elsewhere within the voter registration system.
Questions About Voter Activity
Another category that continues to grow involves registrations showing voting activity after a recorded purge date.
April: 86,276
May: 87,624
An increase of 1,348 records.
This statistic does not necessarily indicate improper voting. It may instead reflect timing differences, administrative processing, or synchronization issues between voter activity and purge records. Nevertheless, the trend warrants further review to ensure that election records accurately reflect voter history and registration status.
Long-Term Inactive Participation
The number of active voters who have not voted in the last two federal elections declined slightly.
April: 1,735,583
May: 1,731,894
However, another statistic moved in the opposite direction.
More than 1.22 million registrations remain associated with voters who have not participated in an election for at least five years.
The NVRA establishes procedures governing how states maintain these registrations while protecting eligible voters from improper removal. Simply not voting does not, by itself, permit cancellation of a registration under federal law, but long-term inactivity remains an important component of lawful list maintenance.
Data Quality Continues to Improve in Some Areas
Several database quality indicators showed modest improvement.
Irregular birthdate records declined slightly.
Registrations using postal addresses remained virtually unchanged.
Only one registrant under age sixteen remained in the database.
No voters were identified without addresses.
These are positive signs that certain data-quality issues continue to receive attention.
Out-of-State Registrations Remain Elevated
One category changed little.
Approximately 246,685 registrations remain associated with individuals believed to have relocated outside New York.
Although this reflects a slight improvement from April, nearly a quarter-million registrations remain in this category.
The NVRA requires states to make reasonable efforts to maintain accurate voter rolls with respect to changes of residence while following statutory safeguards designed to protect eligible voters.
Suspected Deceased Registrations Increased Significantly
Perhaps the second most notable trend involved suspected deceased active registrations.
April: 12,337
May: 16,470
An increase of 4,133 registrations, or approximately 33.5 percent.
The keyword here is suspected.
These records require additional verification before any registration should be cancelled.
Nevertheless, an increase of this size raises reasonable questions that deserve public explanation.
Where Things Stand
The May 2026 Election Integrity Score Sheet reflects both meaningful improvements and significant questions.
Several categories moved in a positive direction, including:
duplicate State Board identification numbers,
irregular birthdate records,
inactive registrations, and
out-of-state registrations.
At the same time, several statistics warrant continued public attention:
Duplicate voter records increased by 285,957.
Suspected deceased active registrations increased by 4,133.
Records showing voting activity after purge dates increased to 87,624.
Nearly 247,000 registrations remain associated with individuals believed to have moved out of state.
More than 1.2 million registered voters have not voted in at least five years.
The most significant question remains the extraordinary increase in duplicate voter records.
Public confidence in elections depends not only on secure elections, but also on transparent election administration.
Accurate voter rolls are not optional.
They are the foundation upon which election integrity is built.
*Data obtained by FOIL from the NYSBOE May 15, 2026 NYSVOTER Snapshot as well as previous monthly snapshots