Quinnipiac has this interesting poll
on the death penalty. The survey was taken in Connecticut, but the numbers run pretty close to the national average. In particular, we see the result of phrasing the question in different ways:
The standard way Gallup has been asking since the 30s:
"Do you favor or oppose the death penalty for persons convicted of murder?" 65-23-12
This is problematic because it implies the respondent must choose a single punishment for all murderers.
The opponents' favorite phrasing, which makes the problem noted above even worse:
"Which punishment do you prefer for people convicted of murder, the death
penalty or life in prison with no chance of parole?" 46-41-14
As strong a proponent of capital punishment as I am, I would answer that question LWOP myself if I were really constrained to a single choice for all murders. Most polls with this phrasing report that a substantial portion of respondents break out of the choices offered and volunteer "it depends" or something like that. No such breakout is reported here. That could be a difference in the way the surveyor follows up if the respondent breaks out of the choices offered. Some polls will simply record the volunteered answer, while others will press for a choice among those given.
And finally, the real question that is actually under discussion:
"Which statement comes closest to your point of view?
(A) All persons convicted of murder should get the death penalty,
(B) No one convicted of murder should get the death penalty, or
(C) Whether or not someone convicted of murder gets the death penalty
should depend on the circumstances of the case." A-7%; B-14%; C-78%; don't know/no answer-1%
Bravo for Quinnipiac! When a pollster finally
asks the real question, 78% of the people favor the basic approach of current law. Another 7% favor making it far more severe than it presently is or has been at any time in modern history, for a total of 85%
in favor overall.
When they ask specifically about Steven Hayes, recently convicted of an exceptionally horrible multiple murder, the result is 76% favor, 18% oppose, 6% don't know. It makes you wonder what the 4% who answered "oppose" to this question but did not answer (B) to the previous question are thinking, but polls often have a few percent of odd answers. The 5% who answered "don't know" to this question but not the previous one may simply be exercising caution in a case where the penalty phase has not been conducted yet.
The poll also finds, unsurprisingly, that very few people would base their vote for governor solely on the death penalty. Again, I would answer that question "no" myself.