Not So Negative?

Buell and Sigelman:

Our measure of negative campaigning is the attack propensity score, which simply states the percentage of all campaign statements that attacked the opponents. Here’s the year-by-year breakdown for each party:

Democrats Republicans
1960 65.6% 45.0%
1964 64.6% 66.3%
1968 57.1% 49.6%
1972 64.1% 49.7%
1976 47.2% 41.6%
1980 54.5% 58.6%
1984 66.0% 38.7%
1988 61.5% 57.5%
1992 59.6% 65.8%
1996 51.8% 48.4%
2000 36.7% 53.2%
2004 71.9% 52.9%
2008 45.6% 37.5%

. . . . Bottom line: Although the 2008 presidential campaign has already emerged as one of the oddest in modern times, its negativity is unusual only by virtue of being less, not more, in evidence.

Take a look at the article; but what this approach doesn't measure is the degree or quality of negativity, merely the crude quantity of negative messages. I wonder if there's any research that gets at that kind of question (that is: surely not all attacks or all lies are created equal). I need to take another look at the book (not handy at the moment), but I don't think that they distinguish between "attack" and "contrast" statements or ads. . . .

UPDATE: Compare with this