StarTribune (PSRA) (11/8, 5/13 in parenthesis):
"Would you favor or oppose amending the Minnesota Constitution to allow marriage only between a man and a woman?"
Favor 48 (39)
Oppose 43 (55)
Don't know/refused 8 (7)
(MoE: ±4.4%)
It's hard to say exactly what happened since May that would result in opposition going from plus sixteen points to minus five, but this certainly helps to illustrate the uncertainties of polling ballot questions, especially this particular ballot question.
[Note: I missed it at first, but the question wording is different from the May poll to this one and that is almost certainly what is driving the results.]
Looking at the cross-tabs the changes come from some expected and some unexpected places. Back in May the 65+ cohort actually opposed the amendment, 51-44, now that same group is in favor of the amendment 70-26. This is a preposterously large swing and may be the root of the dramatic topline swing.
If we look at some previous polls that have been done we find a decidedly mixed picture. SurveyUSA's May survey found a similar level of support among 65+ as the recent Strib survey; 66% supporting the amendment and 27% against. While in PPP's June survey they found 57% support and 34% opposition.
Clearly the 65+ age group is in favor of the amendment, but to what degree? If the recent Strib poll and SurveyUSA are right, opposition to the amendment might be in more trouble then I had previously thought. If, on the other hand, PPP is right, then we are in for the tough fight that we all expect.
More troubling perhaps, this isn't a case of Republicans "coming home," instead the increase in support is coming almost exclusively from Democrats and independents. In the Strib's previous poll Democrats opposed the amendment 71-24, but now that opposition is only 60-35, a stunning 22 point decline. And independents have gone from opposing the amendment 57-34 to supporting it 48-44, a 27 point swing.
There is simply no way we can win this if 35% of Democrats support the amendment. Unifying the DFL on this issue has to be priority number one going forward or we're in big trouble.
---
Below are all the polls so far done on the anti-family amendment (at least all the one's I've got in my spreadsheet, let me know if I've missed any*) with a simple average at the bottom.
Pollster | median date | "No" | "Yes" | Un | margin |
---|---|---|---|---|---|
PSRA | 5/4/2011 | 55 | 39 | 7 | 16 |
SurveyUSA | 5/24/2011 | 40 | 51 | 10 | -11 |
PPP | 5/29/2011 | 47 | 46 | 7 | 1 |
PSRA | 11/3/2011 | 43 | 48 | 8 | -5 |
Ave | 46 | 46 | 8 | 0 |
If we learned anything about polling ballot questions from the Legacy amendment, it's that you need to mirror the actual wording on the ballot to get the best results. This result is probably more accurate.
ReplyDeleteUnfortunately you're absolutely right.
ReplyDelete