Contrary to IHE's report, there's nothing at all surprising about this so-called "finding," given the massive self-selection effect of who takes a survey that includes "bullying" as a topic, and given the massive inflation of the concept to the point that it means nothing. As the first commenter at IHE amusingly puts it: "Does 'hostile and intimidating behavior' include being told no or just being frowned at or simply being challenged by a colleague?" I'm sure in the minds of many it does.
UPDATE: Philosopher Mike Titelbaum (Wisconsin) writes:
Just read your blog post on the UW-Madison survey results and bullying. Having taken that survey myself, I agree with many of the commentators on the IHE page that the notion of bullying in question was underspecified in the survey. I honestly was not entirely clear what I was answering. But I disagree with you that there was a "massive self-selection effect". First, when the survey was sent out, it did not describe itself as including "bullying" as a topic. One had to get well into the survey before that even came up. Second, and more importantly, if the IHE is reporting correctly, then 35% of respondents reported bullying, and the survey had a 59% response rate from faculty. So even if there was self-selection, over 20% of the total faculty reported bullying. While self-selection may have inflated the result from somewhere over 20% to the reported value of 35%, 20% is still a significant number (to the extent that reporting "bullying" is significant in any way).
Given the cultural inflation of "bullying" noted by Professor Collins (in the link, above), I still think this tells us more about cultural training in sensitivity and willingness to take offense than about actual conduct in the real world.