Political Slant of United States Daily Newspapers, 2005 (ICPSR 26242)

Principal Investigator(s): Gentzkow, Matthew, University of Chicago, and National Bureau of Economic Research (NBER); Shapiro, Jesse M., University of Chicago, and National Bureau of Economic Research (NBER)

Summary: The focus of this data collection was media slant in news coverage in the United States in 2005. Automated searches of newspaper articles and congressional records were conducted of 1,000 key phrases addressing issues such as abortion, gun control, taxes, health care, war, the environment, immigration policy, stem cell research, and minorities. A new index of media slant was then constructed that measured the similarity of a news outlet's language to that of the congressional Republican or ... (more info)

• These data are available only to users at ICPSR member institutions. Because you are not logged in, we cannot verify that you will be able to download the data.

WARNING: This study is over 150MB in size and may take several minutes to download on a typical internet connection.

Gentzkow, Matthew, and Jesse M. Shapiro. Political Slant of United States Daily Newspapers, 2005. ICPSR26242-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2009-12-09. doi:10.3886/ICPSR26242.v1

Persistent URL: http://dx.doi.org/10.3886/ICPSR26242.v1

Export Citation:

• RIS (generic format for RefWorks, EndNote, etc.)
• EndNote XML (EndNote X4.0.1 or higher)

Summary:   The focus of this data collection was media slant in news coverage in the United States in 2005. Automated searches of newspaper articles and congressional records were conducted of 1,000 key phrases addressing issues such as abortion, gun control, taxes, health care, war, the environment, immigration policy, stem cell research, and minorities. A new index of media slant was then constructed that measured the similarity of a news outlet's language to that of the congressional Republican or Democrat. To measure news slant, the researchers examined the set of all key phrases used by each congressperson and identified those used much more frequently by one political party than by another. Newspapers were then indexed by the extent to which the use of politically charged phrases in their news coverage resembled the use of the same phrases in the speech of a congressional Democrat or Republican. Part 1, Newspapers File, lists the slant index of 434 city newspapers across the United States and the total number of uses of key phrases for each newspaper. Part 2, News Counts File, indicates the number of news articles in each newspaper using each key phrase. Part 3, Congress File, lists the United States congressperson from each congressional district, his or her political party identification, the slant index, and the total number of uses of all key phrases for this congressperson. Part 4, Congress Counts File, includes information on the number of uses of each key phrase by each member of Congress. Part 5, Phrases File, includes detailed information on the key phrases used in the searches and the total number of uses of each phrase for all Democrats and Republicans in Congress.

Subject Terms:   media coverage, media influence, news media, newspapers, political ideologies, political parties, political partisanship, social issues, United States Congress

Geographic Coverage:   United States

Time Period:  

• 2005

Date of Collection:  

• 2006-06--2009-06

Universe:   English-language daily newspapers with United States circulation in 2005.

Data Types:   administrative records data, aggregate data, event/transaction data

Data Collection Notes:

For more information on sampling, data collection methods, and how the media slant index was constructed, please refer to the published article "What drives media slant? Evidence from U.S. Daily Newspapers" available on the National Bureau of Economic Research Web site.

Part 1 can be linked to Part 2 using the variable NEWSPAPER_ID. Part 3 can be linked to Part 4 with the variable CONGRESS_ID. Parts 2 and 4 can both be linked to Part 5 using variable PHRASE_STUB.

In Part 5, Phrases File, some values in the variable PHRASE are truncated or misspelled.

Some values in variable NAME in Part 3, Congress File, contain diacritical marks.

The data in this collection were generated by computer algorithm from original sources and were produced by the University of Chicago, Chicago, IL.

Sample:   This sample includes 434 English-language daily newspapers available in either ProQuest or NewsLibrary databases, which together represented 74 percent of the total circulation of daily newspapers in the U.S. in 2001.

Data Source:

2005 Congressional Record, downloaded from The Library of Congress Web site.

The NewsLibrary database, downloaded from the The NewsLibrary Web site.

ProQuest Newsstand database, downloaded from the ProQuest Newsstand Database Web site.

Original ICPSR Release:  2009-12-09

• List all 28 variables in this study

• Search the variables in this study

Use any of the notification links to add this study to your RSS feed; you will then receive notification if the study is substantively updated.

• Citations exports are provided above.

• Export Study-level metadata (does not include variable-level metadata)

  • DDI Codebook or DDI Lifecycle

  • Dublin Core

  • MARC21 XML

  • DataCite XML

If you're looking for collection-level metadata rather than an individual metadata record, please visit our Metadata Records page.

• View Usage of this Study