When asked if the doctor of a patient with an incurable disease should be allowed “to assist the patient to commit suicide” at the patient’s request, 51 percent of Americans said it should be permitted. However, when the question was rephrased to ask if the doctor should be allowed “to end the patient’s life by some painless means” at the patient and their family’s request, support shot up to 70 percent.
Polarization across party lines also decreases when “suicide” is not used to describe the practice. The conventional framing solicits support for the practice from 60 percent of Democrats and 41 percent of Republicans, a gap that’s reduced to 71 and 68 percent with softer language.
Support for the question with “suicide” in its phrasing has fluctuated over the past two decades. Whereas it stood virtually at the same level of support as it does today in 1996, support reached a peak of 68 percent in 2001 before gradually falling over the next ten years. Meanwhile, support for the “by some painless means” phrasing has remained over 64 percent throughout that time frame. In the late 1990s, public visibility of physician-assisted suicide, or euthanasia, was high amidst ballot measures legalizing and attempting to repeal the practice in Oregon in 1994 and 1997, and the high-profile trial and arrest of Dr Jack Kevorkian in 1998-99.
The Vermont legislature passed a bill legalizing physician-assisted suicide in May, making it the fourth state to legalize the practice. Washington state and Oregon did so through ballot measures in 1994 and 2008 respectively, and a Montana court ruling made the practice legal in 2008. A bill similar to Vermont’s was defeated in the Maine House last Friday, and a Massachusetts ballot measure was narrowly defeated in the 2012 election.
Even outside of physician-assisted suicide, the language used to communicate end-of-life decisions has been shown to have significant effects. In May, University of Pittsburgh researchers showed that when asked whether CPR should be used on critically ill patients — a measure that only works 10 percent of the time — family members were much less likely to say yes to CPR if the decision was framed as “allow a natural death” (49 percent) rather than “do not resuscitate” (61 percent).
Kumar Ramanathan is an intern at ThinkProgress.