For anyone feeling smugly superior because they don't associate with a party:

Pause, and ask yourself whether there is something special about those labels, or whether the effect probably applies just as strongly when a policy is likely to be favored or disfavored by "scientists" or "nerds" or "Christians" or "rationalists."

And for those who are thinking of simple remedies involving anonymity, I wonder what portion of the "unbiased" assessments involved (conscious or unconscious) speculation about what the opinions of various groups would likely be; I am quite curious if there's anyway we could test this more explicitly.

[edited to add:] I should note that both of the above were me before further reflection - I'm posting lest anyone else stop there.