Alive Enough to Vote

Evaluating Voter Verification Systems Two Decades Later

In the two decades since Congress passed the Help America Vote Act of 2002, the machinery of American election integrity has undergone a profound technological transformation. What began as a mandate for statewide voter databases and basic identity verification has evolved into a complex ecosystem of real-time data matching, involving the Social Security Administration’s Help America Vote Verification (HAVV) system and, increasingly, the Department of Homeland Security’s Systematic Alien Verification for Entitlements (SAVE) program. A close examination of how these tools have performed since the problematic 2005 era—when a GAO report documented Texas officials discovering an attempt to register using 42 deceased individuals’ names—reveals a mixed record of progress, persistent limitations, and urgent questions about how states should deploy verification technology before future elections

HAVV: The Data Backbone

SSA’s HAVV system remains the most established tool in the verification arsenal. When a voter registration applicant lacks a driver’s license, election officials submit the individual’s name, date of birth, and the last four digits of their Social Security Number to SSA. The system returns one of several verification codes, including Code 6, indicating that the data matched SSA records but that the individual is flagged as deceased. This is a critical gatekeeping function: a deceased match means the registration should be halted pending further review.

The HAVV system has improved measurably since 2005. Transaction volumes are now published weekly at ssa.gov/data/havv, providing a degree of transparency that was absent in the early years. States have moved from batch processing to more routine electronic verification. The AAMVA documentation confirms that HAVV returns granular result codes—allowing officials to distinguish between a simple name mismatch, a non-existent SSN, and a deceased individual. In theory, this precision should prevent the kind of undetected fraud that the 2005 Texas case exemplified.

Yet HAVV is not a cure-all. The system only verifies what states submit; it cannot detect registrations that use false information never sent to SSA. Moreover, because HAVV is triggered primarily when applicants lack driver’s licenses, it captures only a subset of the registration stream. And while the system flags potential deceased matches, public weekly data aggregates transactions without publishing state-by-state counts of Code 6 results, meaning outside observers cannot independently audit whether deceased-flag rates are anomalous in specific jurisdictions.

SAVE: The Controversial Expansion

If HAVV represents the steady evolution of identity verification, SAVE represents a more volatile and politically charged expansion into citizenship verification. Originally designed to verify benefit eligibility for immigrants, SAVE was dramatically repurposed beginning in 2025 to audit voter rolls for alleged non-citizen registrations. This expansion has proven deeply problematic.

Texas’s experience is instructive. In April 2026, former Texas Secretary of State Jane Nelson informed USCIS that her office had identified 2,724 potential non-citizens through SAVE—but emphasized that county-level investigations were ongoing and that the data required careful review. Votebeat Texas subsequently reported that SAVE was “mistakenly flagging voters as noncitizens,” and the Brennan Center for Justice warned that the tool’s known inaccuracies posed severe risks of wrongful disenfranchisement. Put simply, SAVE’s error rate is high enough that officials in Texas—hardly a jurisdiction known for lax election oversight—publicly questioned its reliability.

The distinction between HAVV and SAVE is crucial here. HAVV checks identity and life status against SSA records with relatively straightforward matching logic. SAVE, by contrast, attempts to adjudicate citizenship status using DHS databases that were never designed for real-time voter roll maintenance. Data lags, naturalization record delays, and inconsistent state-to-federal data sharing mean SAVE routinely flags naturalized U.S. citizens as aliens. Where HAVV provides clarity, SAVE too often creates confusion.

The Missouri Question

Against this backdrop, the approaching November election demands rigorous scrutiny of how states are using these tools. Missouri has reportedly utilized verification systems in ways that generated striking figures—in one reportedly cited instance, tens of thousands of registrations flagged for verification concerns in a single week. Whether these numbers reflect HAVV deceased matches, SAVE citizenship flags, or some combination of systems, the scale alone underscores the importance of process integrity.

Missouri could learn from both Texas’s caution and California’s model. Texas officials recognized that raw SAVE hits require human review and county-level confirmation before any removal action; California, meanwhile, runs continuous cross-checks between its VoteCal system and state death records from the Department of Public Health, catching discrepancies before they propagate into disputed registrations. Missouri should adopt a similarly multi-layered approach: using HAVV for identity verification, treating SAVE data as a leads list rather than an execution order, and mandating secondary confirmation—such as contact with the voter or cross-reference with vital records—before canceling any registration.

Recommendations for November

With election integrity justifiably a vital concern, states preparing for November should remember three lessons from the post-2005 era. First, verification tools are only as good as the processes surrounding them. An automated flag from HAVV or SAVE should trigger due process, not automatic cancellation. Second, transparency matters. If Missouri or any state is sitting on thousands of flagged registrations, the public deserves clarity about what systems generated those flags and what error rates those systems carry. Third, training remains essential. The 2005 Texas case involved a human or systemic failure that allowed fraudulent registrations to reach the point of submission; modern technology should reduce that risk, but only if local officials understand the difference between a HAVV mismatch and a SAVE false positive.

The system is working better than it did in 2005, but “better” does not mean “flawless.” HAVV has become a reliable identity-verification backbone, while SAVE remains a troubled tool whose misuse could disenfranchise eligible voters. As November approaches, states would do well to rely more heavily on the former while treating the latter with far greater skepticism—ensuring that the pursuit of election integrity does not come at the cost of legitimate voters’ constitutional rights.