Gallup's reporting raises questions

One reason The New York Times’ reputation has sunk in recent years has been its willingness to use headlines to spin stories. “Fake but accurate” comes immediately to mind. Another reason is The Times' habit of placing far down a story important information that goes counter to its headline spin.

I thought about that yesterday when I visited The Gallup Organization’s website and read its report comparing poll results of blacks’ and whites’ responses to a range of questions related to Katrina.

Gallup headlined:

Blacks Blast Bush for Katrina Response
Most believe racism was responsible for delays in providing relief

Its lead paragraphs support the headlines:

Whites and blacks have sharply differing reactions to the federal government's response to Hurricane Katrina, with blacks more likely than whites to believe that racial bias was a factor in slowing the government's response, and blacks especially critical of President Bush's performance.

Aside from Bush, whites and blacks have similar perspectives on how various entities handled themselves in the Hurricane Katrina disaster. Most whites and blacks agree that FEMA, state and local Louisiana authorities, and New Orleans residents did a poor job initially, but that in recent days each has been doing a good job.

Some questions are in order here.

Why did Gallup decide to headline: Blacks Blast Bush for Katrina Response? Haven’t blacks, by percentages sometimes in the 90s, been blasting Bush for years for just about everything he’s said or done? What’s headline-worthy about them blasting him for his Katrina response?

And what about the following information which isn't reported until the eighth paragraph of Gallup’s story:

The percentage of blacks approving of the job Bush is doing as president is 14% today, little different from the 15% recorded in August. Similarly, the percentage of blacks holding a favorable view of the Republican Party is now 16%, versus 18% earlier this year.

So in the aftermath of Katrina, Gallup’s own polling finds no statistically significant difference from pre-Katrina levels in blacks’ approval of the President or their favorable view of the Republican Party. Those are important findings given all that's been said these past few weeks.

Why didn’t Gallup think them worth a headline or at least a mention in one of the first few paragraphs?

The answer’s obvious, isn’t it? “No change post- Katrina in blacks’ approval of Bush or favorable rating of Republican Party” just doesn’t go with:

Blacks Blast Bush for Katrina Response
Most believe racism was responsible for delays in providing relief

In a similar way, when Gallup told people in its second paragraph that:

Aside from Bush, whites and blacks have similar perspectives on how various entities handled themselves in the Hurricane Katrina disaster

it had the problem of what to do with some of its findings that contradicted its "whites and blacks have similar perspectives" claim. Findings such as:

While only 17% of blacks have an unfavorable opinion of New Orleans Mayor Ray Nagin, 38% of whites view him unfavorably.

Three-quarters of blacks (77%), compared with only 44% of whites, consider the post-hurricane looters in New Orleans to have been mostly desperate people trying to keep themselves alive.

Half of whites and only 16% of blacks (viewed the looters) as mostly criminals taking advantage of the situation.

Gallup made no reference to those findings in its lead paragraphs.

I’m going to send Dr. Frank Newport, Editor-in-Chief of The Gallup Organization, a link to this post and invite him to respond. I’ll publish in full his response.

Like most Americans, I respect The Gallup Organization and count on it for objective presentation and interpretation of data scientifically obtained.

I hope those headlines and lead paragraphs aren’t a sign it’s drifting in the direction of The New York Times.

I’d hate to wake up one morning, go to Gallup’s site, and see the headline: Our latest poll results fake but accurate.