Brown University economics professor Jesse Shapiro co-authored a recently released study on the use of partisan language in Congressional speech from 1873 to 2009, which suggests that American politics are more polarized than ever before.
Shapiro and his partners in the study developed and used a machine-learning algorithm, which, based on phrases used by politicians in their speeches, tried to guess the speakers party affiliation. It was 54 to 55 percent accurate until 1994, when its accuracy spiked.
“Partisanship was low and roughly constant from 1873 to the early 1990s, then increased dramatically in subsequent years,” the study’s authors wrote. “Beginning with the congressional election of 1994, partisanship turned sharply upward, with the probability of guessing [a speaker’s political party] correctly based on a one-minute speech climbing to 83 percent by the 110th session (2007-09).”
The study’s authors were surprised by the
immediate and dramatic increase, which they attribute to the increased use of
political marketing tactics built around consultants, focus groups and polls.
The authors isolate the 1994 Contract with American, wherein Republicans
detailed their legislative objectives for the first 100 days of that session,
as a key turning point in partisan speech.
The study contends that partisan language
flows from politicians into the media and public discourse.