Washington Post headline:
"Which is more taboo: The Confederate flag or the rainbow flag?"
Oh, boy.
The piece is written by two political scientists and reports on survey data, done before the church shootings in South Carolina or the Supreme Court's marriage equality decision, showing who was and was not offended by the two flags.
Most Americans said they were at least “slightly offended” by the Confederate flag (13% said slightly, 20% said moderately, 12% said very, and 15% said extremely offended). About 39% said that they were “not offended at all.” Fewer people took offense at gay pride flag — 56% said they were not offended at all — as of November of last year.
This is a case where data journalism desperately needs context to save it from total inanity. As interesting and relevant as it might be to know how many Americans find these things offensive, and how those views change over time, a nothing-but-the-numbers write-up of survey data, promoted as stand-alone journalism, is ... lacking.
The thing is, there's no straight comparison to be made between the two flags. The implication is that it's a racism vs. homophobia question, but things are a lot more complicated than that. It's more like how many people are offended by the existence of out LGBT people vs. how many both are offended by racism and connect the Confederate flag to racism. The rainbow flag is just a simpler symbol as a relatively recent creation with a straightforward association with gay pride. By contrast, the Confederate flag has a Civil War history and a history, growing out of that Civil War history, as a symbol of resistance to integration and the civil rights movement. Both those histories are racist, but the long "heritage not hate" PR campaign, however false, complicates the flag's status as a direct proxy for racism. How the different historical associations with the Confederate flag break down in survey responses isn't clear, and it's not made any more clear in this write-up.
It's a great example of how statistics don't always tell us anything. But I don't think that's what the Post thought it was doing here.