I have some doubts about whether the study really measures "bias" in a simple sense. I have no doubt that the did things technically right -- I vaguely know the lead author, Tim Groseclose, from when we were both at MIT, and he's a smart guy (Tim himself is very conseravative politically, for what that's worth).

The problem is this: essentially what they do is say that if a highly partisan Democrat (defined as one who votes the way the party leadership wants all the time) tends to mention the Brookings Institution a lot in his speeches, but not the Heritage Foundation, then if CBS News also cites Brookings but not Heritage, then we conclude that CBS is biased towards Democrats. I'm sure the UCLA team did a good job technically in measuing those things and doing good stats on them.

But are think tank citations really a good proxy? I'm not so sure. One issue is that think tanks get cited a lot more on some issues than others. You'll see a lot of competing citations of studies when it comes to say, regulation of banks, but not so much when it comes to school prayer or confirmation of presidnetial appointees, yet those sorts of votes tend to be overrepresented in the indexes used to calculate those party loyalty scores.

One reason the WSJ comes out as having the most liberal bias -- yes, that's what they find -- is, I suspect, that it has the greatest focus on business news, which is where you find the greatest concentration of think tank work, especially on the liberal side. You see some conservative think thanks addressing social issues, but few liberal ones (compare Heritage and Brookings, for example); in the liberal area, social issues tend to be focussed on by activist groups, not think tanks.

Likewise, you find a much higher percentage of conservative think thanks outside of DC. Those non-DC outfits frequently get cited by Congressmen from their states, but much less so by national media outlets. Thus you could have a bias that reflects laziness more than editorial stance.

Anyway, I look forward to reading it, but I worry they carried their bridge a bit too far.