One brief follow-up. The piece pivots off of a statement by political scientist Joseph Heim, who says that negative ads can be effective for undecided voters or earlier in the campaign. I actually agree with this somewhat. As I’ve noted before, campaign advertising is likely to have a larger impact when the candidates are unfamiliar, which in turn makes voters less certain who to vote for. Presidential primaries, which is what Heim was talking about, often feature unfamiliar candidates. Ads in this GOP primary may have mattered, at least by my back-of-the-envelope calculations from before the Florida primary.
The issue, I think, is that we don’t know that negative ads necessarily have a larger impact than positive ads even in races with unfamiliar candidates, lots of undecided voters, etc. As Phil Arena has noted, the decision to “go negative” is hardly independent of where the candidate is in the polls. It’s not clear what’s the cart and what’s the horse—ads or poll numbers.
Another point I made in the interview, although this didn’t make it into the story, is that in real-life campaigns feature all kinds of ads—purely positive, purely negative, contrast—and typically these ads are being aired at the same time. It’s hard to find a controlled setting—except in the proverbial laboratory—where you can isolate the effects of negative ads.