During the manual recount and examination of Florida ballots from the 2000 Presidential election, late-night comedians and editorial cartoonists had a field day with jokes about aging Florida voters. One cartoon from the Minneapolis Star-Tribune went as far as to call some voters “confused, simple-instructions-challenged Florida retirees.”

But it was the infamous butterfly ballot—a paper with candidates’ names listed on both sides and punch holes in the middle—that caught Jason Karlawish’s attention. As an assistant professor in the Division of Geriatrics at Penn’s School of Medicine and a fellow in the Center for Bioethics, Karlawish has focused his research on ethical issues in the care of patients with dementia—defined as those experiencing a decline in two or more cognitive functions, like visual and special reasoning. “This may be why a ballot like this caused such problems,” said Karlawish, showing a slide of a butterfly ballot during a presentation at the Center for Bioethics Oct. 5.

With his Dementia Voting Project, Karlawish seeks to identify and address legal, ethical and medical issues regarding the voting rights and abilities of people with dementia. With support from the Greenwald Foundation, Karlawish has assembled a team of experts from law, ethics, government, neurology, geriatrics, psychiatry and long-term-care to understand what it means to be capable of voting and how caregivers should help people with dementia exercise the right to vote—a fundamental right of all U.S. citizens.

Caregivers in nursing homes and assisted living facilities—places where patients with dementia often live—should have guidelines for determining their patients’ ability to vote, Karlawish told his lunchtime audience. In his research, he determined that in 31 nursing homes and 20 assisted living facilities in Philadelphia, the way caregivers decide competency varies widely. Some make a subjective assessment of the cognitive skills of patients, while others ask a series of election-related questions to assess patients’ competency, a practice that Karlawish dislikes. “It harkens back to literacy tests in the South in the ’50s and ’60s,” he said.

In his research, Karlawish also interviewed 33 people with mild to severe Alzheimer’s to assess their understanding of the voting process, and found that about half didn’t understand the nature of voting, while half could compare candidates running for office. He believes that adults who lack the capacity to understand the nature and effect of voting may legitimately be precluded from the process. “Simply being able to choose a candidate does not mean you fully understand the nature and effect of voting,” he said. “Merely saying ‘I want to vote’ doesn’t translate into knowing what voting is.”