Policy Brief on Using Databases to Keep Eligible Voters Off the Rolls

[PDF]

Summary

* Ill-conceived state policies concerning new statewide voter registration databases are keeping eligible voters off of the rolls, through no fault of their own.
* Database matching can be unreliable.
* States can ensure that voter registration lists are as complete and accurate as possible while still safeguarding voters’ rights.
* States should adopt flexible matching standards for new registrants.
* Most states have protective procedures in the event no match can be found.
* Protecting legitimate voters also requires common-sense technological safeguards.

Ill-conceived state policies concerning new statewide voter registration databases are keeping eligible voters off of the rolls, through no fault of their own. Across the country, states are implementing the Help America Vote Act of 2002 (“HAVA”), creating new statewide computerized databases of registered voters. The technology has the potential to improve the registration process substantially. However, it also has the potential to be quite dangerous – in particular, when officials unduly rely on the ability to “match” information from one source to another. A few outlier states have now created a new illegal precondition for registration: the state’s ability to match information on voter registration forms to information in other government databases, such as the state’s motor vehicles database or the federal Social Security system. In these states, if the government cannot find a “match” for information on the form, the applicant will not be registered and will not be able to cast a valid ballot.

Database matching can be unreliable. Unfortunately, the matching process is often fraught with error. All large databases contain mistakes – typos or transposed fields, for example, that would prevent records from matching even when they represent the same person. Also, databases record information inconsistently, which makes it even more difficult to find proper matches: “William” may not match “Will” or “Billy”; a name may be spelled “Mohammed” or “Muhammad”; a maiden name may not match a married name. If the state must be able to produce a match before a voter can be registered – and therefore, eligible to vote – huge numbers of eligible citizens may be mistakenly disenfranchised through no fault of their own. In Washington State, for example, one woman’s birth date was entered into the system as “1976” instead of “1975” (the year written on her registration form), and when no matching record could be found, her registration form was rejected. Another form was rejected because the voter was listed with a maiden name in one source and a married name in another.

Nor are these isolated incidents. To the contrary, the error rates have been remarkable. A sample run in New York City in 2004, for example, showed that if the right to vote were conditioned on a proper match, up to 20% of new voter registrations would have been rejected solely because of data entry errors. Similar “matching” error rates of 20–30% were discovered in Washington State. And the Social Security Administration has reported a 28.5% failed match rate nationwide.

States can ensure that voter registration lists are as complete and accurate as possible while still safeguarding voters’ rights. States need not fall prey to the “no match, no vote” failures above. Federal law asks states to try to match registration information to other government databases in order to validate the unique number assigned to every individual in the statewide registration system. However, the law allows states to set flexible standards for determining when a match is found. And federal law requires states to register an eligible voter even if the state cannot locate matching information elsewhere.

States should adopt flexible matching standards for new registrants. States seeking to make their lists as accurate as possible without jeopardizing eligible voters should adopt flexible standards that improve the reliability of the matching process. These flexible standards would account for typos, nicknames, and other inconsistencies among imperfect databases. Some states have indicated that they plan to pursue such a path. In a Brennan Center survey conducted in the fall of 2005, for example, Arkansas indicated that it would attempt to find matches for driver’s license information based on individual review of a range of reasonable, substantially similar possibilities. Unfortunately, in comparing Social Security information, states may not have such an option: the Social Security Administration appears to have set a national standard – a character-by-character computerized “exact match” – that is especially prone to mistakes. Keeping the matching criteria flexible and subject to human review would better compensate for common errors.

Most states have protective procedures in the event no match can be found. As mentioned above, a few outlier states illegally disenfranchise applicants when the state cannot produce matching information in a government database. Most states, however, do not attach such unwarranted consequences to a procedure that will frequently fail. Indeed, the usual course is to place an eligible applicant on the statewide list of registered voters even if matching information cannot be found.

For some of these registrants, federal law asks for additional validation of the registrant’s identity: citizens who are registering for the first time in the state, and doing so by mail (without, for example, registering before an election official), must either have their registration information confirmed by another government source or must show some sort of identification before voting a regular ballot. Most states will flag these registrants on the pollbooks so that any issues can be resolved easily on Election Day. In each such case, however, the applicants’ registration status is not in jeopardy.

Protecting legitimate voters also requires common-sense technological safeguards. Many states are currently constructing the large statewide voter registration databases for the first time. In addition to comprising the official list of voters – and thereby determining whether any individual is ultimately able to vote – the databases will also contain a substantial amount of private personal information. States must therefore implement common-sense technological protections for these enormous systems, such as requiring that a log of all database transactions be maintained, in order to track and remedy improper access. These databases must also be protected by layers of access and authorization to ensure that only authorized transactions are made and only by authorized people.

THE WORK OF THE BRENNAN CENTER

* National. In March of 2006, the Brennan Center released the first national survey of states’ policies and practices in registering voters – and matching their registration information – using the new statewide voter databases. The survey remains the most comprehensive collection of practices using the new computerized registration systems. In addition, this report included detailed policy recommendations based on best practices in the states and comparative research from other fields.

* Washington. The Brennan Center recently won the first lawsuit in the country confronting a “no match, no vote” policy. In August 2006, a federal court blocked the implementation of a Washington State law that would have barred citizens from voting unless the Secretary of State first succeeded in matching information from voter registration forms with records kept in other government databases.

* California, Maryland, Pennsylvania, New Jersey. Throughout 2006, the Brennan Center has been working with state officials to ensure that the registration databases become helpful tools and not barriers to voting. Thanks in part to Brennan Center legal analysis and advocacy, Maryland and Pennsylvania have reversed their “no match, no vote” policies, and California and New Jersey appear to be well on the way to doing the same.

Last Updated September 2006