The fact that disability agencies were not included in the 2012 litigation might help explain why they fared worse than the county assistance offices in registering people to vote. But it's likely not the only reason.
"My impression is that most disability services are too many layers down from the state's accountability," Regal adds. He notes that Just Harvest contracts with the state to help provide public assistance. But he couldn't say for sure whether there was a provision in the contract that requires he participate in voter-registration efforts.
"When DPW or the Department of Health enters into a contract with a private company to outsource some of its functions, they're supposed to delegate to the contractor its NVRA duties," says Ben Geffen, a lawyer at the Public Interest Law Center in Philadelphia. "Our understanding is that compliance is spotty at best."
The Department of Public Welfare, which runs county assistance offices, did not respond to numerous phone calls and emails seeking comment over a nearly week-long period.
Jonathan Marks, commissioner for the Bureau of Commissions, Elections and Legislation — and who helps oversee data collection on state compliance with the NVRA — says much of the concern over the state's compliance with the NVRA can be attributed to data-collection systems that don't tell the full story.
"I think we've demonstrated ... that a lot of this is about tracking — and has been all along," Marks says, noting that it's possible voter registrations by public-assistance agencies have been historically undercounted.
For one, Marks explains, the state has long relied on generic voter-registration forms, which he says can be traced to specific public-assistance agencies only if they are funneled directly by those agencies to the county elections office.
And because an increasing number of people can apply for services like Medicaid and food stamps without ever showing up at a welfare office, many of those registration applications have gone out by mail, potentially bypassing the state's ability to count them.
Since 2009, Marks says, the state has moved toward coded forms, which are more easily traced to specific agencies. But they're not automatically distributed (the agencies have to request them), likely meaning the data is still distorted. "In some counties, old forms were still circulating out there," he says.
And asked why the rates of disability service providers registering people to vote tend to hover in the triple digits, he said "there's some overlap" between county assistance offices and disability providers. One person might get services from both and therefore have already been asked about voter registration, for instance, while others might be children or otherwise ineligible to vote.
Asked about those who fall through the cracks, Marks says, "I don't think we've dismissed the possibility that people are being missed, but I don't think that's a significant proportion."
O'Hanlon acknowledges Marks could be right, but he counters that if the state did offer more people voting-registration opportunities than the data show, "they don't appear to be in a position to prove that."
O'Hanlon's concern over the data released this past June isn't just over the comparatively weak registrations collected by disability agencies. He also points to statistics that show a large number of what he calls "lost voters": people who were presumably offered an opportunity to register, since they did not decline to be registered, but somehow never made it onto the rolls.
Under the NVRA, if you do not decline to be registered to vote, you are still essentially treated as if you want to register.
For instance, of the 2.7 million people who either said "yes" or did not decline when asked by county assistance offices if they would like to register to vote, only 47,028 were actually registered to vote. "When I look at the lost voters, it makes me think maybe we're only seeing a fraction of the improvement we should be seeing," O'Hanlon says.
Among disability service providers, meanwhile, 8,039 did not decline to be registered to vote last year. But that generated just 514 registrations.
For his part, Marks explains that many of those lost voters did not explicitly say they wanted to register to vote. And many of them were likely never asked whether they wanted to vote by a human being, so it's possible they just ignored the question and were not interested in registering, Marks adds.
But Richard Weishaupt, a lawyer at Community Legal Services in Philadelphia, says the state shouldn't just assume people aren't registering because they don't want to.
"People go to those offices often under a great deal of stress," he says, "and they're not thinking primarily about their voter registration at the moment," making it more important to affirmatively help people register.
Regal agrees — though he notes that hoping for good political will toward people on public assistance might be a lofty expectation. "If you could register online, like you can pay your taxes online and like you can do your banking [...] then it would be easy and we wouldn't have these problems," he says.
"These are the people who went to court to protect voter ID. How eager is the state government to help people register to vote?"
