Shortly before the 2016 election, Mary Saucedo sat down with her husband of 52 years, Gus, to fill out her absentee ballot. Mary, a resident of Manchester, New Hampshire, is currently 95 years old and legally blind. She has voted consistently since 1944. Federal and state law both permit Gus to help his wife cast her vote, but Mary wanted to sign the ballot herself, so he guided her hand to the affidavit and she affixed her signature. Gus mailed her ballot to the Manchester clerk’s office. She had every reason to believe her vote would be counted.
It wasn’t. On Election Day, a local official compared the signature on Mary’s ballot to the one on her absentee ballot affidavit, a document she’d signed to attest that her disability prevented her from voting in person. The official determined that the two signatures didn’t match and tossed out her ballot. There was no opportunity for appeal. Mary didn’t know she’d been disenfranchised until the American Civil Liberties Union of New Hampshire informed her in 2017.
That election cycle, an estimated 275 New Hampshire voters suffered the same fate, losing their right to vote because an official with no expertise in handwriting analysis decided that two signatures didn’t match. In the 2012 election, as many as 370 voters in the state were disenfranchised through this scheme. It’s not clear that a single one of these voters knew what happened to their ballots—at least not until the ACLU of New Hampshire told them. The organization is now suing the state in federal court for flagrant violation of federal law and the U.S. Constitution. It will very likely win in a rout.
Why are so many New Hampshire voters disqualified on the basis of their handwriting? The answer to that question is oddly elusive. Voters’ signatures are analyzed by “moderators” at each polling place. The number of ballots rejected by moderators due to signature “mismatch” varies wildly from town to town. In 2016, all such rejected ballots came from just 26 percent of New Hampshire’s 318 polling places; the remaining 74 percent of polls rejected zero ballots for mismatch. It thus appears that a minority of overzealous moderators are responsible for this disenfranchisement.
These moderators’ analysis of ballot signatures is shockingly amateurish: They briefly compare the two names, decide whether they match, and either accept or reject the ballot based on their subjective determination. There are no objective standards to guide the process. Dr. Linton A. Mohammed, a forensic document examiner, testified to the ACLU that effective signature comparison—which is difficult even for experts—requires 10 signature samples “at a minimum” to account for variability. Moderators only have two: the signature on the affidavit requesting a ballot and the signature on the ballot itself. With this limited sample, Mohammed explained, any analyst—and especially one with no training—is “likely to erroneously conclude that two signatures from the same voter do not match.”
Voters are not informed directly when their ballots are rejected due to a signature mismatch. Instead, municipal clerks add their names to a website days or weeks after the election. Then, 90 days after the election, the information is scrubbed from the internet. The state does virtually nothing to educate voters about the existence of this site. Voters also have no way to resolve the issue; they are not allowed to prove their identities after the fact. Once a vote is discarded by a moderator, it’s lost forever.
No other state employs a scheme as draconian as New Hampshire’s, though some have tried.
Making matters worse, New Hampshire only allows residents to vote absentee if they have a legitimate impediment to voting at the polls, such as a disability. For this reason, a disproportionate number of absentee voters are disabled. The ACLU found that many of those whose votes were tossed out because of a supposed signature mismatch lived in assisted-living facilities, including a number of retired service members in veterans’ homes. Presumably, age and disability prevented them from affixing perfectly consistent signatures to their ballots. Sadly, a number of them died following the 2016 election. The last vote they cast was not counted.
In light of this problem, the New Hampshire Legislature recently adjusted its voting laws to clarify that disabled voters may receive help filling out their ballots so long as an assistant submits a statement declaring he helped out. Yet, even under this new law, a moderator can still reject the ballot for signature mismatch. And of course, individuals who cannot obtain effective assistance, or have no disability, are not helped by this tweak at all. The Secretary of State’s Office even acknowledged to the legislature that this putative fix does not “allay” the New Hampshire ACLU’s lawsuit.
That suit alleges, quite plausibly, that the state’s mismatch scheme is illegal under the Americans with Disabilities Act, which requires states to enact voting procedures that grant equal access to disabled residents. It also claims that the scheme violates procedural due process under the 14th Amendment by depriving voters of any recourse to cure the loss of their vote. And it argues that New Hampshire infringes upon its residents’ fundamental right to vote as protected by the Equal Protection Clause. Any law that severely burdens the franchise, as this scheme clearly does for hundreds of voters each election, must survive strict scrutiny to pass constitutional muster. In this case, New Hampshire plainly cannot put forth a legitimate interest to justify its stringent mismatch rule aside from a generalized (and empirically unsupported) fear of voter fraud, which cannot satisfy heightened scrutiny.
No other state employs a scheme as draconian as New Hampshire’s, though some have tried. Judges have struck down mismatch laws in Florida, California, and Illinois that do not give voters notice or an opportunity for remedy. A federal judge found that Florida’s rule would’ve disqualified as many as 23,000 ballots in the 2016 election; the ACLU estimates that California’s disenfranchised 40,000 voters that same year. Plenty of other states also compare signatures, but they all provide notice and a chance to fix any purported mismatch. New Hampshire’s stringent rule seems to be singular.
The ACLU filed its first complaint in this case in May. Since then, the Republican-controlled legislature has had ample opportunity to fix the glaring legal infirmities in its mismatch scheme. Instead, legislators have forged ahead with voter suppression laws targeting lower-income residents and college students. So, on Monday, the ACLU filed for summary judgment, asking the court to rule in its favor without a trial. It hopes to win the case in time for the 2018 midterms so people like Mary Saucedo can ensure their votes are counted. Saucedo testified that she plans to cast an absentee ballot once again in the next election—as long as “I’m still here.”
In 2018, we’re turning our attention to the cornerstone of our democracy: the right to vote. Slate’s journalists will be offering expanded coverage of gerrymandering and voter suppression, investigating key legislative battles and court cases on the state and federal level. They’ll also offer new tools to help readers understand how the electoral sausage gets made.
Fund Our Work