Attention conservation notice: Self-promoting a dry academic comment on a bad book about bias in the US media.
It seems that the Perspective on Politics review symposium on Groseclose's Left Turn, organized by Henry Farrell, is finally out. This includes a contribution by Justin Gross, Andrew Gelman and your humble narrator, focusing on the methodological problems, which are, to my mind, crippling. (Of the other contributions, the only one I've had a chance to read is Brendan Nyhan's, which, unsurprisingly, I like.)
There's no abstract, so I'll just quote our opening.
In Left Turn, Groseclose concludes that, in a world without media bias, the average American voter would be positioned at around 25 on a 0--100 scale, where 0 is a right-wing Republican and 100 is a left-wing Democrat. In this world, a balanced media might include some TV networks promoting the view that abortion should be illegal under all circumstances and subject to criminal penalties, whereas others might merely hold that Roe v. Wade is unconstitutional; some media outlets might support outright discrimination against gays whereas others might be neutral on civil unions but oppose gay marriage; and on general politics there might be some newspapers that endorse hard-right Republican candidates (0 on Groseclose's 0--100 scale) whereas those on the left would endorse positions near those currently held by Senator Olympia Snowe. But instead of this, Groseclose must endure a world where he estimates the average voter is around 50, with all that follows from this, and he attributes this difference to media bias.
Groseclose sets up this stirring climax by formulating and estimating three models. The first model, from Groseclose and Milyo (2005), is an ideal-point model that puts media organizations and members of Congress on that 100 point scale, based on how often they refer to various research and advocacy organizations. The second infers the political positions of voters in different districts based on how many of them voted for Obama, and the estimated positions of their members of Congress. The third model, new to Left Turn, is a causal model of how media influence the positions of voters. Groseclose's claims about what our country would look like if it weren't for media bias thus rest on a multi-stage analysis. (Figure 1 depicts his estimation strategy.) He estimates latent quantities, such as how Americans would vote if their views were not distorted by the media, in terms of other latent, estimated quantities, such as the political location of media organizations.
Groseclose's estimation strategy. Variables in boxes are observable; variables in ellipses are latent, and estimated. (The causal network implied by Groseclose's models is subtly different from this inferential network.) PQ and SQ stand for Groseclose's "political quotient" and "slant quotient", respectively.
You can read the rest at the journal, with nice formatting, or a preprint. That last link also carries you to the appendices on technical issues we couldn't fit into in the main article. Even the appendices leave out something which struck me when I read the book, that the models make no sense as strategic actors rationally maximizing utility. This doesn't make them any worse to my mind, but is the kind of thing the people who write and edit the Quarterly Journal of Economics purport to find troubling...
(As usual, I'm not speaking for the statistics department or CMU here, and everything in this post which goes beyond the article isn't the fault of my co-authors, etc.)
Please comment on the article here: Three-Toed Sloth