Republican and Conservative political analysis and commentary

Slate's Jack Shafer has an interesting article about the various catch phrases favored by Republicans and by Democrats at the national level, and how a new academic study by a couple of economists shows how those catch phrases can be used to spot media political bias. Writes Shafer:

It's common knowledge that political consultants prescribe to their clients specific phrases to use when discussing hot topics. Republican consultant Frank Luntz, Gentzkow and Shapiro write, urged Republicans to refer to "personal accounts," not "private accounts," when talking about changes in Social Security. Similarly, Luntz suggested the phrase "tax relief" to repel the Democrats' complaint about "tax breaks." George Lakoff has written a whole book advising Democrats on what words to use to frame the debate in their favor.

Shafer's article looks at a new academic study of hundreds of such catch phrases - and how often the media uses one or the other while reporting on politics and government.

The paper, "What Drives Media Slant?," is a working paper on media bias by University of Chicago scholars Matthew Gentzkow and Jesse M. Shapiro, who list a few hundred two- and three-word phrases that Gentzkow and Shapiro say can be uniquely identified with either Republican or Democratic members of Congress.

Shafer explains the research:

To capture the politically loaded terms, the authors used computers to scrape the 2005 Congressional Record. Floor speeches by members of both parties in both houses were analyzed, and based on frequency of use, key phrases were assigned to one party or the other. Next, the authors measured the "slant" or bias of hundreds of newspapers by noting the frequency with which these loaded phrases appear in news reports during 2005. Essentially, the more often a newspaper echoes Republican or Democratic phrases, the more Republican or Democratic its bias. Gentzkow and Shapiro do much, much more with the data, mashing it up against campaign donations by zip code to show how market forces help determine news content, and flinging other socio-political-economic data around to create a model to predict the level of bias they expect to find at individual papers.

You can read the abstract or download the whole study here.

This caught my eye: the authors of the study found that "the average newspaper's language is similar to that of a left-of-center member of Congress." The study results say much of this is due to newspapers seeking to cater to the political slant of their audience.