Trends in Political Values and Core Attitudes, 1987-2009

Data Archive > U.S. Surveys > General Population > National > Pew Research Center > Summary


The values study by the Pew Research Center for the People & the Press is an effort to better understand the nature of American politics by tracking a broad range of beliefs and values that shape public opinion and ultimately influence voting behavior. The project began in 1987, and has been updated 13 times over the past 22 years. The values study asks respondents whether they agree or disagree with a series of approximately eighty statements covering core beliefs about government, business, religion and several other topics.

(From the Pew Research Center for the People & the Press report "Independents Take Center Stage in Obama Era: Trends in Political Values and Core Attitudes: 1987-2009," available here)

Data File
Cases: 32,570
Variables: 170
Weight Variable: WEIGHT
Data Collection
Date Collected: Data from the fourteen waves of this study were collected during the following months: May 1987, May 1988, February 1989, April to May 1990, October to Novemeber 1991, June 1992, May 1993, July 1994, November 1997, October 1999, August 2002, August 2003, January 2007, and April 2009.
Funded By
The Pew Research Center for the People & the Press
Collection Procedures
In-person interviews (1987-1990) and telephone surveys (1991-2009)
Sampling Procedures
The target population was comprised of adults age 18 or older, residing in households in the United States. In-person interviews were conducted from 1987 to 1990, and telephone surveys were conducted thereafter. The telephone sample was provided by Survey Sampling International, LLC (SSI). Using the list-assisted random-digit-dialing methodology, the list was created to give each number an equal chance of being selected. After selection, numbers matching those of businesses were purged.

*Note: These sampling procedures, used in the 2007 and 2009 surveys, are the most recent sets and are different from those utilized in previous surveys.

Up to 10 attempts were made to contact each number in the sample list over a staggered number of days to increase the likelihood of contacting potential respondents. Once contact was established, interviewers asked to speak with the youngest male resident or, if this person was not present, the oldest female resident. This procedure produces results that closely mirror the population in age and gender.
Principal Investigators
The Pew Research Center for the People & the Press
Related Publications
The Pew Research Center for the People & the Press Survey Report: Independents Take Center Stage in Obama Era: Trends in Political Values and Core Attitudes: 1987-2009
Note 1: RELIG Details
The original question wording for the variable RELIG prompted interviewers to code those who provided the following responses as Protestants: "Baptist, Episcopal, Jehovah's Witness, Lutheran, Methodist, Presbyterian, Episcopalian, Pentecostal, Church of Christ, etc."
Note 2: INCOME Coding
INCOME is a variable on which interviewers assigned respondents to one of four broad income categories (low, lower middle, upper middle, or upper) based on their more detailed income estimates (see INCOME87, INCOME89, INCOME91, INCOME94, or INCOME07 for these estimates). Income levels that led interviewers to assign respondents to different income categories on the variable INCOME varied by year, as specified below:

INCOME87 (survey years 1987, 1988 and 1990):
1. Respondents with annual household incomes of $0 to $9,999 were classified as belonging to the "low" income bracket.
2. Respondents with annual household incomes of $10,000 to $24,999 were classified as belonging to the "lower middle" income bracket.
3. Respondents with annual household incomes of $25,000 to $39,999 were classified as belonging to the "upper middle" income bracket.
4. Respondents with annual household incomes of $40,000 or greater were classified as belonging to the "upper" income bracket.

INCOME89 (survey year 1989):
1. Respondents with annual household incomes of $0 to $14,999 were classified as belonging to the "low" income bracket.
2. Respondents with annual household incomes of $15,000 to $24,999 were classified as belonging to the "lower middle" income bracket.
3. Respondents with annual household incomes of $25,000 to $39,999 were classified as belonging to the "upper middle" income bracket.
4. Respondents with annual household incomes of $40,000 or greater were classified as belonging to the "upper" income bracket.

INCOME91 (survey years 1991, 1992 and 1993):
1. Respondents with annual household incomes of $0 to $14,999 were classified as belonging to the "low" income bracket.
2. Respondents with annual household incomes of $15,000 to $29,999 were classified as belonging to the "lower middle" income bracket.
3. Respondents with annual household incomes of $30,000 to $49,999 were classified as belonging to the "upper middle" income bracket.
4. Respondents with annual household incomes of $50,000 or greater were classified as belonging to the "upper" income bracket.

Classification from the INCOME94 variable (survey years 1994, 1997, 1999, 2002 and 2003) varied by year in the following way:
1. Respondents with total family incomes of $0 to $19,999 were classified as belonging to the "low" income bracket.
2. In survey years 1994, 1997 and 1999, respondents with total family incomes of $20,000 to $29,999 were classified as belonging to the "lower middle" income bracket. However, in survey years 2002 and 2003, respondents with total family incomes of $20,000 to $20,999 as well as those with total family incomes of $30,000 to $39,999 were classified as belonging to the "lower middle" income bracket.
3. In survey years 1994, 1997 and 1999, respondents with total family incomes of $30,000 to $74,999 were classified as belonging to the "upper middle" income bracket. However, in survey years 2002 and 2003, only respondents with total family incomes of $40,000 to $74,999 were classified as belonging to the "upper middle" income bracket.
4. Respondents with total family incomes of $75,000 or greater were classified as belonging to the "upper" income bracket.

INCOME07 (survey years 2007 and 2009):
1. Respondents with total family incomes of $0 to $19,999 were classified as belonging to the "low" income bracket.
2. Respondents with total family incomes of $20,000 to $39,999 were classified as belonging to the "lower middle" income bracket.
3. Respondents with total family incomes of $40,000 to $74,999 were classified as belonging to the "upper middle" income bracket.
4. Respondents with total family incomes of $75,000 or greater were classified as belonging to the "upper" income bracket.
Note 3: INCOME91 Details
Respondents' income estimates on the variable INCOME91 was derived from a series of contingency questions, as shown below:

1. Is your total annual household income before taxes $20,000 or more or is it less than $20,000?
2A. IF LESS THAN $20,000: Is it over or under $15,000?
2B. IF LESS THAN $15,000: Is it over or under $10,000?
3A. IF MORE THAN $20,000: Is it over or under $30,000?
3B. IF MORE THAN $30,000: Is it over or under $50,000?
Note 4: FIPS Codes
To translate FIPS (Federal Information Professing Standards) codes, refer to the Federal Information Processing Standards Publication 6-4 at the following link:
http://www.itl.nist.gov/fipspubs/co-codes/states.txt

The variable FIPS presents each respondent's five-digit FIPS code. The first two digits of this code denote their state and the final three digits of this code denote their county. Values on the variable STATE2 denote each respondent's state and the two-digit code associated with that state, while values on the variable COUNTY denote the respondent's county, with reference to the respondent's state.
Note 5: MSA Codes
To translate MSA (Metropolitan Statistical Area) codes for survey years 1987, 1988 and 1989, refer to the following U.S. Census publication:
http://www.census.gov/population/www/metroareas/lists/historical/83mfips.txt

To translate MSA (Metropolitan Statistical Area) codes for survey years 1990, 1991 and 1992, refer to the following U.S. Census publication:
http://www.census.gov/population/www/metroareas/lists/historical/90mfips.txt

To translate MSA (Metropolitan Statistical Area) codes for survey years 1993, 1994 and 1997, refer to the following U.S. Census publication:
http://www.census.gov/population/www/metroareas/lists/historical/93mfips.txt

To translate MSA (Metropolitan Statistical Area) codes for survey years 1999, 2002 and 2003, refer to the following U.S. Census publication:
http://www.census.gov/population/www/metroareas/lists/historical/99mfips.txt

To translate MSA (Metropolitan Statistical Area) codes for survey years 2007 and 2009, refer to the Office of Management and Budget's (OMB) Bulletin No. 08-01 at the following link:
http://www.whitehouse.gov/omb/bulletins/fy2008/b08-01.pdf