Message Testing

Message Testing


J Street reader JL emailed yesterday to make an interesting point:

I think that once political campaigns make decisions about how to characterize their opponents, they grow to believe these characterizations, and, as a result, are slow to change course when the characterizations don’t take hold with the electorate. This explains the months Hillary spent trying to paint Obama as being “just words” before finally switching to an approach that paints him as an elitist black radical who hates America. I think the Democrats are running full wind into this same mistake when it comes to McCain. The WSJ has a story this morning showing that the GOP has national approval ratings in the 20s, but that McCain is stubbornly in the mid-40s. A strategy that tries to paint McCain as four more years of Bush, which is what Howard Dean is pushing from the DNC, just isn’t going to work, because most people have a view of McCain from 2000 and beyond as a Bush antagonist. This is going to stick even if it turns out that he shares an Iraq strategy with the President (he’s actually more hawkish). Instead of trying to sell him as four more years of Bush, the Democrats should paint him as a flip flopping hypocrite who isn’t nearly as pure as he tries to come off. That will expose him as just a regular pol, with limited economic competence.

Good point! Although, I actually think JL is too high an information voter to be particularly representative. But it turns out, there’s actually some data on this. NBC/WSJ just ran a somewhat odd poll [PDF] that tested three negative messages against each candidate to see which ones resonated most. Turns out both the “he’s too close to Bush” and “he’s a flip-flopper with now real convictions” test well, the former (pace JL) just edging out the latter. You can see the results presented in a highly parsable fashion here.

Also of note, the attacks against Obama that have been getting the most press of late, resonate the least with voters. Surprising.

Ad Policy