Who Is Really Worried?
David Broder informs us that he has spoken to Republicans who are very worried about November. I'll take him at his word. There are people who should be worried.
The impression these Republicans had is that support for GOP candidates had nose-dived this summer — in part because of the chaos conveyed by the daily televised scenes of destruction in Iraq and Lebanon and in part because of the dismal reputation built by the Republican Congress that is home to many of the endangered GOP candidates.
It may be that the cease-fire in the Israel-Lebanon war and the shift of focus to the terrorist plot thwarted last week by the British will change the political environment. But Republicans were deeply worried as August began.
I had dinner one night with a group of Ohio Republicans, all with many years of experience in state politics and none directly engaged in this year's gubernatorial race. One of them said, "I'm afraid this could be another 1982," a year when recession pushed unemployment to 15 percent and cost the Republicans the governorship. Another said, "I'd settle right now for another 1982. I'm afraid it will be another 1974," the year of the Watergate election, when Democrats swept everything in sight.
Ohio may be particularly vulnerable because the economy in parts of the state where the auto industry remains vital has been hurt by layoffs, and because a series of scandals has left retiring Gov. Bob Taft with approval ratings in the teens. But similar concerns are voiced across the Midwest.
Here's the thing though. At the end of the article, Broder mis-characterizes an AP-Ipsos poll. His words don't match the poll results.
That prediction is supported by the AP-Ipsos poll released last week. It showed a 33 percent job approval score for Bush and a 29 percent job approval score for the Republican Congress. On a test of voter preference for the midterm congressional elections, Democrats had a staggering 18-point lead, 55 percent to 37 percent. You can see why Republicans are worried.
Ok, bad numbers indeed. But Broder identifies it as the "Republican Congress". That is not what the poll asked. The poll asked about Congress. Period. No party affiliation. The exact question:
Overall, do you approve, disapprove or have mixed feelings about the way Congress is handling its job?
Note that party is not mentioned in any way. This question captures dissatisfaction with both parties equally, there is no way to attribute it to either party. Here is the other thing that is very skewed in the poll. Initial identification of respondents party:
Republican ………………………………… 27
Democrat…………………………………… 35
Independent ………………………………. 24
None of these…………………………….. 13
Not sure…………………………………… 1
Clarified after additional questions:
Strongly Republican …………………….. 16
Moderately Republican…………………. 23
Definitely Independent/neither……….. 7
Moderately Democrat …………………… 32
Strongly Democrat……………………….. 21
Refused/not sure……………………….. 1
Total Republican ………………………. 39
Total Democrat …………………………. 53
This sample is almost laughably skewed. There is an 8% skew just in the initial sample. The clarification skews it even further to 14%. That's almost enough to account for the result. So who is actually worried to be spinning this poll this furiously?





