GameOver22 said:
Why would it be the best measure of ideology? I'm kind of confused about that. If extreme votes are disproportionately represented on roll-calls, you are going to get biased estimates of true ideology because you won't pick up more moderate issues. |
Because it's the issues without cross contamination.
These are the issues less likely to have "Vote trading".
Also, by focusing on extreme issues. It sort of makes bias ring clearer. By picking the extremely polarized votes... you isolate the positions that are extremely polarized.
In otherwords, the ones in which reporting definitly should not be taking a side.
It's not exagerrating bias' in the media... if there was any criticism of it... you would say it's not making enough differentiation in the "middle."
As a 48 and a 52 would be far more apart then you would expect.
What the Gloscuse study more or less does... is find out how most media outlets stand on the most extreme polarizing and divided issues.