The Washington Post published a minor but disgustingly lazy article called
How Donations Depict Donors. It's more proof (if you need it) that the press too often lazily outsource analytical work, then report it as fact without subjecting the analysis to even basic standards.
By starting with incomplete data and refusing to add multivariant data to the mix, the net effect of this "analysis" is to misleadingly stereotype a candidates' supporters by wealth, class, and lifestyle.
In order to write the article, the WaPo got some trashy demographics firm to analyze individual donor data to draw donor profiles. Presumably the data is from Q3, though they don't say so outright. They then shoehorn supporters into predefined categories based on their home zip code. For example, among Dean supporters are the "Bohemian Mix" who like Audi A4s and shop at Banana Republic. Kerry nabs the "Young Digerati" and Bush takes the "Winner's Circle." Blah blah blah.
But the worst part of the article is lack of context. They don't bother to mention that only individual donors who give candidates more than $200 in an election cycle are required to be identified by name and address. So whereas (for example) over 50% of Dean's supporters cannot be profiled, they are invisible, not heard. How many of those AFSCME and SEUI members shop at Banana Republic, anyway?
Other data should have been referenced for a more complete picture. For example, data on average donation amounts (e.g. $74 for Dean, $280 for Bush), breadth of support, and number of overall donations would help readers better understand the candidates' donors -- hopefully without the broad brush of lifestyle stereotyping. I've never shopped at Ann Taylor or Banana Republic, and I never will.
I doubt excellent WaPo journalists like Mike Allen are quaking in their boots -- the likes of these two "journalists" aren't exactly nipping at anyone's heels. Still, the editors let this profile in laziness pass under their noses without a smell test, and should be ashamed.