Dueling Polls--Or Just Dueling Headlines?

From USAToday:
Poll sees a boost for Bush, Iraq war

From CBS:
Poll: Most still think Iraq war a mistake

Curious, is it not, that two polls in the wake of al-Zarqawi's death could seemingly reach such different conclusions? One finds a majority still saying it was "a mistake" while another sees a boost for the war's support? How does that happen?

Well, it happens because of the headline chosen, which is my own personal beef with "polls." They are rhorshach tests, allowing people to see almost anything in their results (a practice in which I will now engage). For example, USAToday led with the fact that their poll found that 47% believe things are going well in Iraq, up from 38% in March. The survey also showed Bush's approval rating going up to 38% from 36% earlier this month and an all-time low of 31% in May.

CNN headlined the fact that their poll showed 55 percent of respondents believing the U.S.-led invasion of Iraq in 2003 was an error -- a figure unchanged from an April survey.

The fact that Bush's approval rating went up two points doesn't strike me as news, especially when the next quote in the article has Stuart Rothenberg, a political analyst, saying "good news makes people feel better." Duh. And especially when there is no discussion of whether a 2-point jump is even statistically significant. Nevertheless, all of the 2%s and 12%'s aside regarding people's short-term emotions about the war, the fact that 55% continue to think the invasion was an error--an unchanged figure no matter the death of al-Zarqawi--strikes me as the real piece of news here, and as the pivotal point of pain for the political fortunes of GWB and the Republican Party. That unchanged figure ought to signal loud and clear that there has been a loss of trust between the American people and their President. That short-term gains in the war will obviously "make people feel better" but will not alter their belief that the Iraq invasion was a bad decision with negative consequences across the board. If I were the political advisor to the POTUS or to a GOP member of congress, I would take those numbers to heart in my electioneering. It would be tempting to focus on the "but he gained 2 points since last month!" or "12% more people think the war is going well!" but a truly smart politician will recognize the shifting sands of short-term electorate emotion when she sees them, and opt instead to address the larger issues: trust between a country and its President, acknowledgement of previous error or at least capricious unilateral action, resisting the urge to convince people that it really wasn't an error, or blindly supporting the administration.

Obviously my hope is that the Republicans will NOT heed this advice and will lose spectacularly in the mid-term elections, but the message is the same for Democrats, on a whole host of issues: stop grasping for the numbers you WANT to see and start addressing the numbers that are there.

I charge $300 an hour if anyone needs me. ;)