| Looking at Survey USA's trendlines for Culver since he took office, I noticed that Culver's disapproval number is basically unchanged this summer, but his approval number has dropped significantly from 42 percent in June and 44 percent in July to 36 percent in late August.
Before anyone panics, note that Survey USA's approval numbers for Culver tend to run low compared to other pollsters. In early July, the poll commissioned by The Iowa Republican blog found Culver's approve/disapprove numbers to be 53 percent/41 percent. Later the same month, Hill Research Consultants' poll for the Iowa First Foundation found Culver's favorability at 52 percent. (The Iowa First Foundation did not release the governor's approval number from that poll, but you better believe they would have if the number had been in the 30s or even the low 40s.) Meanwhile, Survey USA pegged Culver's approval at 44 percent on July 20.
Survey USA's numbers for Obama, Harkin and Grassley are also noticeably down in the latest poll. Obama is at a record low in Iowa. Grassley's approval of 54 percent is the lowest Survey USA has found in at least four years. I couldn't find a similar graph for Harkin's numbers, but it's been a very long time since I can remember seeing his approval rating below 50.
Of course, it's possible that the recession and the health care debate have affected Iowans' view of all political leaders. Still, I would like to see these numbers confirmed by some other pollster. Even with the best sampling techniques, approximately 1 in 20 polls is wrong just by chance.
I also agree with Steve Singiser that if Culver were this unpopular in Iowa, Democrat Curt Hanson would not have won yesterday's special election in Iowa House district 90 (a swing district). The Republicans ran at least two television ads linking Hanson to Culver (see here and here).
I'm looking forward to the next Selzer and Associates poll for the Des Moines Register, which probably will come later this month or in early October.
Share any relevant thoughts in this thread.
UPDATE: Bleeding Heartland user ragbrai08 noticed this:
The sample composition:
D-28
R-34
NP-35
Obviously, if considering registered adults, the D/R fractions should be reversed. However, back in July, they used:
D-38
R-24
NP-34
so the Dem fraction was too high back then. Conclusion? Perhaps the July survey was a bit too favorable, and perhaps this one a bit too unfavorable.
In an off-year election, turnout among no-party voters is likely to be lower. Even with a demoralized Democratic base, I'd be surprised if Iowa's electorate next year was comprised of 28 percent Democrats and 34 percent Republicans. Right now Democrats have a voter registration advantage of about 100,000 over Republicans. |
