> These high-empathy students were also more likely to be amused by reports that students protesting the speech had injured a bystander sympathetic to the speaker.

> one gauges people’s level of “empathic concern” by asking them how strongly they agree or disagree with a series of seven statements such as “I often have tender, concerned feelings for people less fortunate than me.”

It seems like a problematic way to measure empathy. I don't think it would be surprising that people who claim to have a lot of empathy in surveys, or in general conversations, would demonstrate a lack of empathy in their actions.

My tongue-in-cheek version of the abstract would be something like "...We argue that, in practice, the _self-reported_ experience of empathy ... can actually exacerbate political polarization."