Elections: party identification in the 2004 election.

Winneg, Kenneth ; Jamieson, Kathleen Hall

Background

Since Belknap and Campbell tied the concept of party identification to political behavior in 1951 (Belknap and Campbell 1951), party affiliation has played a central role in explanations of individual political behavior in the United States. Indeed, as Nie, Verba, and Petrocik noted in 1979 (p. 47), in the 1950s and 1960s, party identification "was the central thread running through interpretations of American politics" where it was considered "a stable characteristic of the individual: it was likely to be inherited, it was likely to remain steady throughout the citizen's political life, and it was likely to grow in strength during that lifetime."

The drop in the number of citizens reporting a strong tie to party, a drop that occurred between the mid-1960s and mid-1970s and leveled off in the late-1970s and early-1980s, did not change the fact that most continued to identify with a party. But the change in partisan affiliation did focus scholars on its change from one election to another. Abramson, Aldrich, and Rohde spotted the phenomenon on which we focus here in 1994, when they observed that "over the past forty years, the balance between the two parties has favored Democrats by a range of about 55/45 to about 60/40. While the results from the last four presidential election years still fall within that range, they show a clear shift toward the Republicans" (p. 224).

Both the 2004 National Annenberg Election Survey (NAES) and surveys by the Pew Research Center for the People and the Press show that the gap in self-identified party affiliation between Democrats and Republicans is closing. However, because Republicans continue to have higher turnout than Democrats, this difference did not produce a net Democratic advantage on Election Day 2004 when, for the first time since modern exit polling began in 1976, the number of self-identified Republicans who voted equaled the number of people saying they were Democrats (37 percent each; National Election Poll, Pew Research Center for the People and the Press 2005). As Kerry-Edwards's pollster Mark Mellman observed at the 2004 Annenberg Election Debriefing, "In the 70s and 80s, Democrats on Election Day had 15-point margins. By the time we got to the 80s, those were 2 and 3 and 4-point margins. It is right to say that today this was the first election where the exit polls showed parity" (Jamieson 2003). In 2000, the NAES and the U.S. National Election Study (NES) observed a narrowing of the gap (Johnston, Hagen, and Jamieson 2004, 42-43). While this narrowing was picked up in pre-election polls, it had not yet occurred in Election Day exit polls.

Research Questions and Methodology

This Election Day parity invites the questions that we address in this article. How stable was party affiliation during the course of the entire 2004 presidential campaign? Was Republican self-identification consistently rising as the campaign progressed, or was there a noticeable level of variability across the campaign? Also, how does 2004 party identification compare with 2000? Were any changes across 2004 simply the continuation of a pattern evident four years earlier? If there is a narrowing gap between Republicans and Democrats, how does that change the demographic and geographic makeup of the parties? Finally, what impact, if any, has the shift had on the number of registered voters who call themselves Independents?

Our analysis goes beyond the extant research which has a focus on party changes from campaign to campaign over time. Instead, the NAES examines changes that occur within the course of an election year.

Using a rolling cross-sectional design (Romer et al. 2004) created for it by Richard Johnston who had pioneered its use in Canadian elections, the 2004 NAES interviewed 81,422 adults by telephone from October 7, 2003 through November 16, 2004; 67,777 registered voters were part of the 2004 sample. The 2000 NAES was conducted by telephone from December 14, 1999 through January 19, 2001 among 58,373 adults, of which 46,697 were registered voters (for an analysis of 2000 NAES data see Johnston, Hagen, and Jamieson 2004). (1) The question the survey employs for party identification asked: "Generally speaking, do you usually think of yourself as a Republican, a Democrat, an Independent, or something else?" Interviewers recorded verbatim responses from those who said "something else." The question wording used by the NAES is similar to the NES. The only difference is that instead of using the phrase, "or something else," the NES uses the phrase, "or what." NAES interviewers asked the party identification question before the vote intention question, with several questions in between. However, research suggests that question order has no significant impact on party identification (McAllister and Wattenberg 1995).

Results

The results we present regarding party identification do not include "leaners": Independents or "others" who are asked whether they lean more to the Democratic or Republican parties. We present only the root question in our findings because the inclusion of leaners tends to contribute to greater volatility in party identification.

In 2004, NAES results showed that 31.8 percent of registered voters called themselves Republicans and 34.6 percent said they were Democrats, a Democratic edge of 2.8 percentage points. The margin of sampling error on these findings was just over one third of one percentage point in either direction. These findings are similar to the Pew Research Center's aggregate survey numbers on party identification from 2004, which showed a 33 percent to 30 percent Democratic party edge (Pew Research Center 2005). The nearly three-point edge for the Democrats in party identification from the NAES is one point narrower than the advantage our survey showed during the 2000 election campaign period. In 2000, the NAES showed a Democratic edge of 33.7 percent to 29.9 percent, a difference of 3.8 percent (see Table 1). The margin of sampling error for the 2000 findings was less than one half of one percentage point in either direction.

In our analysis for this article, we focus on party identification among registered voters for the entire field period of NAES 2000 and NAES 2004. However, NAES numbers previously published in Johnston, Hagen, and Jamieson's The 2000 Presidential Election and the Foundations of Party Politics (2004) show aggregate party identification from July to Election Day, 2000 among all respondents. For comparative purposes, we present a similar table from the 2004 NAES. Table 2 shows a narrowing of the gap among all respondents.

Demographic Breakdowns of Party Identification

The breakdown by demographic and geographic variables is based on registered voters for the entire field periods of 2000 and 2004. Republican gains since the 2000 election were made across many key demographic variables. The greatest increases were recorded among white evangelical Protestants. In 2000, 42 percent called themselves Republicans, and 25 percent called themselves Democrats, a 17 percentage point Republican advantage. In 2004, the margin increased to 25 points: 48 percent said they were Republicans and 23 percent said they were Democrats. We saw slight increases in both Republican and Democratic party identification among Catholics; they remained a predominantly Democratic group (see Table 3).

Larger demographic gains were seen among those with only a high school education or less (26 percent to 30 percent). Also, Republican party identification grew significantly among both married women and Southerners (31 percent to 35 percent, respectively).

Women are still substantially more Democratic; men are more likely to self-identify as Republicans. In 2000, 28.0 percent of women called themselves Republicans and 38.8 percent said they were Democrats, an advantage of 10.8 percent. In 2004, that advantage declined by more than one percentage point to 9.5 percent. In this past election, 30.1 percent of women said they were Republicans and 39.6 percent said they were Democrats.

The Republican advantage for men grew slightly in 2004 from 4.1 percentage points to 4.9 percentage points. In 2000, 32.1 percent of men said they were Republicans and 28.0 percent said they were Democrats. In 2004, 33.8 percent of men said they were Republicans and 28.9 percent said they were Democrats. Table 4 shows changes from 2000 to 2004 among various population groups.

Time-Series Analysis of 2004 Party Identification

One advantage of the NAES's methodology is that it allows researchers to conduct a time-series analysis to determine at what point during the campaign period changes occur and what events took place at or prior to these changes. Figure 1 uses time-series analysis to show the variability in party identification during the 2004 campaign.

[FIGURE 1 OMITTED]

Figure 1 shows changes in party identification from October 29, 2003 to November 10, 2004 based on a regression analysis of the daily means for each identification, Republican, Democrat, and Independent ("other" and "don't know" are excluded). The curves are the predicted regression lines from each analysis. The regression shows a slight but steady linear incline of Democrats over the course of the primary and general election campaigns. The movement that Figure 1 shows in party identification, especially for Republicans, is evidence that party allegiance is not a constant, but ebbs and flows as the impact of events and campaigns is felt. Methodologically, this change lends support to the view that it is unwise to include party identification as a variable used in weighting or sample balancing the data.

The Republican line begins above the Democratic line in the early months of the Democratic primary campaign. At about the time of the New Hampshire primary (January 27), the lines intersect and the Republican line dips, reaching its nadir in mid spring. While no causality can be firmly established, a number of events may have contributed to this dip in Republican affiliation. There was negative news about the Iraq war: the U.S. suffered more casualties in the Iraq war in April than in any previous month of the war (Torres 2004; Zoroya 2004), and the Abu Ghraib prison scandal broke in late April. The 9/11 Commission was in session, March through May, and President Bush and Vice President Cheney jointly testified behind closed doors at the end of April. Gasoline prices topped $2.00 nationally for the first time, and Kerry had just launched (on May 4) a $25 million advertising campaign in the key battleground states. While NAES data showed Republican affiliation temporarily declining, the proportion of those who called themselves Independents was increasing, but then began to fall as the Republican party identification began to rise at the end of May. NAES data also showed Republican affiliation increasing steadily throughout the summer, including the time when Sen. Kerry introduced Sen. John Edwards as his running mate in early July, through both the Democratic convention in late July and the Republican convention in late August and early September, while the Swift Boat Veterans for Truth were running their ads attacking Kerry in August, and while the Bush campaign was attacking Kerry on national security issues in late August and beyond. The increase in Republican party identification increased through Election Day, reaching parity with Democratic party identification in mid-September.

Changes in Party Identification by State

Because of the NAES's large sample size, we are able to break down party identification for most states. (2) In 2004, we collected sufficient sample sizes in forty-two of forty-eight states where we conducted interviews (the NAES does not conduct interviews in Alaska or Hawaii). We set a cutoff of 330 cases (5.39 percent margin of sampling error) for inclusion in the analysis. The states where we deemed sample size to be too small were Delaware, North Dakota, Rhode Island, South Dakota, Vermont, Wyoming, and the District of Columbia. Additionally, Idaho, Montana, Nevada, New Hampshire, and New Mexico each had a sufficient sample size in 2004, but an insufficient sample size in 2000, so we only show the 2004 numbers for these states.

We found that the most Republican of the states were: Nebraska (45 percent), Utah (44 percent), Kansas (43 percent), Mississippi (39 percent), Idaho (38 percent), and Montana (38 percent).

The most Democratic states in 2004 were Maryland (44 percent), West Virginia (43 percent), Kentucky (41 percent), New York (41 percent), Louisiana (40 percent), and Pennsylvania (40 percent). Pennsylvania was the only state in the top tier that significantly increased in the size of self-identified Democrats from 2000 to 2004 (37 percent to 40 percent). And while Louisiana remained among the top Democratic states in 2004, it saw a significant decline in the proportion of self-identified Democrats from 46 percent to 40 percent. Oklahoma, one of the more Democratic states in 2000, dropped from 44 percent to 39 percent.

It is not all bad news for the Democrats, however. In 2004, there were significant increases in five states in the number of self-identified Democrats in 2004: Wisconsin (+6 points), Michigan (+5 points), New Jersey (+5 points), South Carolina (+5 points)-the South is not completely lost--and Washington State (+5 points).

The states showing the largest gains in self-identified Republicans in 2004 include Maine (+8 points), Alabama (+7 points), Arkansas (+7 points), Mississippi (+7 points), and Oregon (+6 points). There were states that declined in Republican identification, but those drops were statistically insignificant.

The NAES also found a drop among self-identified Independents in several states. The New England states hold the highest proportion of self-identified Independents. In 2004, four out of the top five self-identified Independent states were in New England: Massachusetts, New Hampshire, Connecticut, and Maine. (3) In both 2004 and 2000, Massachusetts reported the greatest percentage of Independents. Forty-seven percent of Massachusetts registered voters in 2004 and 48 percent in 2000 called themselves Independent rather than Democrats or Republicans. We see this plurality of Independents in only two other states in 2004: Connecticut (41 percent Independents, 30 percent Democrats, and 22 percent Republicans) and Iowa (37 percent Independents, 30 percent Democrats, and 28 percent Republicans).

In states with large Republican gains, not all were at the expense of Democrats. More often it was the case that Independents began crossing to the Republican party. Maine and Mississippi showed significant gains in Republican identification, but this increase was at the expense of Independents rather than Democrats. Similarly, in Florida, Kentucky, Minnesota, and Wisconsin, we saw increases in Republican party identification and no decreases in Democratic party identification. Instead, there were declines in those who called themselves Independents (see Table 5).

Conclusion

In summary, in 2004, NAES data show that during the 2004 presidential election campaign, party affiliation fluctuated and was not entirely stable. As the campaign progressed, both Democratic and Republican party self-identification showed a noticeable level of variability and by the end of the campaign, party identification was moving in favor of the Republicans, but the Democratic party retained its advantage, in aggregate.

While the 2004 NAES data show a narrowing of the gap between Democratic and Republican party identification, the changes continued a pattern evident in 2000 in both the NAES and the 2000 NES telephone component. We observed this trend in analyzing 2000 NAES data among registered voters going back to the beginning of the primary season and from NAES data from July 2000 to Election Day.

There was some change in the demographic makeup of the parties when comparing 2004 to 2000, with the biggest changes observed among white evangelical Protestants. The Republican party made slight gains among men and women. However, the Democratic party retained its edge among women. We observed larger increases in Republican party identification in those with lower education levels.

Geographically, Republican party identification strength grew most in the South. A state-by-state analysis of states with sufficient sample size showed that the Republican party made significant gains in party identification in some southern states, but also made such gains in a northeastern state, Maine, and a northwestern state, Oregon. The Democratic party made gains in identification in a handful of states around the country. In some states where Republican identification increased, it was at the expense of Independents rather than Democrats.

Finally, New England states continue to be the bastion of self-identified Independents. However, NAES data show a decline in self-identified Independents from 2000.

Appendix

Methodology

The National Annenberg Election Survey (NAES) is a survey conducted each presidential election by the Annenberg Public Policy Center of the University of Pennsylvania. The 2004 National Annenberg Election Survey is based on telephone interviews which began October 7, 2003 and concluded on November 16, 2004.

The sample of telephone exchanges called was randomly selected by a computer from a complete list of thousands of active residential exchanges across the country. Within each exchange, random digits were added to form a complete telephone number, thus permitting access to both listed and unlisted numbers. Within each household, one adult was designated by a random procedure to be the respondent for the survey. The interviewing is conducted by Schulman, Ronca, & Bucuvalas, Inc. The results have been weighted to take account of household size and number of telephone lines into the residence and to adjust for variation in the sample relating to geographic region, sex, race, age, and education.

This report deals with two sets of interviews. The 2004 campaign interviewing was conducted from October 7, 2003 through November 16, 2004, reaching 67,777 registered voters (81,422 adults). The 2000 campaign interviewing began on December 14, 1999 and continued through January 19, 2001, reaching 46,697 registered voters (58,373 adults).

In theory, in nineteen out of twenty cases, results for all registered voters in 2004 will differ by just over one third of one percentage point, up or down, from what would have been obtained by interviewing all American adults. For smaller subgroups, the margin of sampling error would be higher. For 2000, the margin of sampling error would be less than one half of one percentage point, up or down. For smaller groups, the margin of sampling error would be higher. The findings for all men and all women in 2004 would be subject to sampling error of one half of one percentage point, up or down. In 2000, the potential sampling error for men and women would be about six tenths of one percentage point from what would have been obtained by interviewing all American adults.

In addition to sampling error, the practical difficulties of conducting any survey of public opinion may introduce other sources of error into the poll. Variations in the wording and order of questions, for example, may lead to somewhat different results.

References

Abramson, Paul R., John H. Aldrich, and David W. Rohde. 1994. Change and continuity in the 1992 elections. Washington, DC: CQ Press.

Belknap, George, and Angus Campbell. 1951. Political party identification and attitudes toward foreign policy. Public Opinion Quarterly 13(4): 601-23.

Jamieson, Kathleen H. 2005. Introduction. In Electing the president, 2004: The insiders' view, edited by Kathleen H. Jamieson. Philadelphia: University of Pennsylvania Press.

Johnston, Richard, Michael Hagen, and Kathleen H. Jamieson. 2004. The 2000 presidential election and the foundations of party politics. Cambridge, UK: Cambridge University Press.

McAllister, Ian, and Martin P. Wattenberg. 1993. Measuring levels of party identification: Does question order matter? Public Opinion Quarterly 59(2): 259-68.

Nie, Norman H., Sidney Verba, and John R. Petrocik. 1979. The changing American voter. Cambridge, MA: Harvard University Press.

Pew Research Center for the People and the Press. 2005. National security more linked with partisan affiliation: Politics and values in a 51%-48% nation. Retrieved March 28, 2005, from http://people-press.org/reports/display.php3?ReportID=236.

Romer, Daniel, Kate Kenski, Paul Waldman, Christopher Adasiewicz, and Kathleen H. Jamieson. 2004. Capturing campaign dynamics: The national Annenberg election survey. New York: Oxford University Press.

Torres, Luc. 2004. Iraq conflict: April: The worst month yet. Guardian, May 1. Retrieved April 7, 2005, from http://www.guardian.co.uk/international/story/ 0,,1207461,00.html.

Zoroya, Gregg. 2004. April is Iraq's deadliest month. USA Today, April 29. Retrieved April 7, 2005, from http://www.usatoday.com/news/world/iraq/ 2004-04-29-iraq-deadliest-month_x.htm.

AUTHORS' NOTE: Adam Clymer, political director of the National Annenberg Election Survey, and Dan Romer, PhD, associate director of the Annenberg Public Policy Center, also contributed to this article.

(1.) See Appendix for a description of the full methodology.

(2.) The state analyses do not represent uniquely drawn state samples, but rather show results in a given state from national rolling cross-sectional samples (NAES 2004 and NAES 2000).

(3.) Vermont and Rhode Island had insufficient sample sizes for this analysis in both 2004 and 2000.

Kenneth Winneg is managing director of the National Annenberg Election Survey. Prior to joining the Annenberg Public Policy Center, he was a vice president at Penn, Schoen, and Berland in New York. Before working at Penn, Schoen, and Berland, he was senior research manager at Chilton Research Services, responsible for managing the ABC News-Washington Post surveys.

Kathleen Hall Jamieson is the Elizabeth Ware Packard Professor of Communication at the Annenberg School for Communication, and Walter and Leonore Annenberg Director of the Annenberg Public Policy Center at the University of Pennsylvania. She directs the National Annenberg Election Survey.

TABLE 1
National Self-Reported Party Identification

Q: Generally speaking, do you usually think of yourself as a
Republican, a Democrat, an Independent, or something
else?

 2004 NAES (a) 2000 NAES (a)

Republican 31.8% 29.9%
Democrat 34.6% 33.7%
Independent (c) 25.2% 26.7%

(a) Asked of 67,777 registered voters between October 7, 2003 and
November 16, 2004.

(b) Asked of 46,697 registered voters between December 14, 1999 and
January 19, 2001.

(c) Independents are respondents who indicate they are "Independents."
This number excludes those who give a response of "something else,"
"don't know," or refused to answer.

TABLE 2
Comparisons to the Party Identification Figures in Johnston, Hagen, and
Jamieson's The 2000 Presidential Election and the Foundations of Party
Politics

Q: Generally speaking, do you usually think of yourself as a
Republican, a Democrat, an Independent, or something else?

 2004 NAES (b) 2000 NAES (a)

Republican 31.1% 27.8%
Democrat 32.4% 30.9%
Independent (c) 36.4% 41.3%

(a) Asked of 36,876 registered voters between July 1, 2000 and
November 6, 2000.

(b) Asked of 36,912 registered voters between July 1, 2004 and
November 1, 2004.

(c) Independents include respondents who respond as
"Independents"/"something else"/"don't know"/"refused."

TABLE 3
Party Identification among White Evangelical Protestants

Q: Generally speaking, do you usually think of yourself as a
Republican, a Democrat, an Independent, or something else?

 Republican Democrat Independent (c)

 2000 2004 2000 200 2000 2004

White evangelical 42% 48% 25% 23% 23% 21%
Protestants

Catholics 28% 29% 36% 37% 28% 26%

(a) Asked of 16,931 registered voters classified as white evangelical
Protestants and 11,359 registered voters classified as Catholics
between December 14, 1999 and January 19, 2001.

(b) Asked of 20,093 registered voters classified as white evangelical
Protestants and 16,491 registered voters classified as Catholics
between October 7, 2003 and November 16, 2004.

(c) Independents are respondents who indicate they are "Independents."
This number excludes those who give a response of "something else,"
"don't know," or refused to answer.

TABLE 4
Party Identification by Key Demographic Groups--2004 versus 2000 among
Registered Voters

 Republican
Generally Speaking, Do You
Usually Think of Yourself as a ... 2000 (a) 2004 (b)

Total 29.9% 31.8%

Men 32.1% 33.8%
Women 28.0% 30.1%

High school or less 26% 30%
Some college 32% 34%
College degree or more 34% 33%
 Men high school or less 27% 31%
 Women high school or less 26% 28%
 Men some college 34% 34%
 Women some college 30% 33%
 Men college degree or more 38% 37%
 Women college degree or more 30% 30%

18-29 years old 29% 28%
30-44 years old 32% 35%
45-64 years old 28% 31%
65 and over 30% 32%
 Men 18-29 32% 30%
 Women 18-29 27% 25%
 Men 30-44 35% 37%
 Women 30-44 29% 33%
 Men 45-64 30% 33%
 Women 45-64 26% 29%
 Men 65 and over 31% 33%
 Women 65 and over 29% 32%

Household income below $35,000 24% 25%
$35,000 to less than $75,000 33% 33%
$75,000 and over 37% 38%
 Men below $35,000 25% 26%
 Women below $35,000 23% 24%
 Men $35,000 to $75,000 34% 35%
 Women $35,000 to $75,000 31% 32%
 Men $75,000 and over 39% 40%
 Women $75,000 and over 35% 37%

Whites 33% 36%
African Americans 5% 5%
Hispanic/Latinos 22% 23%
Asians 23% 25%
Native Americans 20% 22%
 White men 35% 38%
 White women 32% 35%
 African American men 6% 7%
 African American women 5% 4%
 Hispanic/Latino men 23% 26%
 Hispanic/Latino women 21% 20%

Union household 23% 24%
 Union men 25% 25%
 Union women 21% 22%
 Married 33% 36%
 Single 24% 24%
 Married men 35% 38%
 Married women 31% 35%
 Single men 26% 25%
 Single women 23% 22%

Northeast 27% 28%
Midwest 29% 31%
South 31% 35%
West 33% 33%
 Northeast men 28% 31%
 Northeast women 26% 24%
 Midwest men 31% 33%
 Midwest women 27% 29%
 Southern men 33% 36%
 Southern women 29% 34%
 Western men 36% 35%
 Western women 30% 31%

Urban 25% 26%
Suburban 32% 34%
Rural 31% 34%
 Urban men 28% 28%
 Urban women 22% 24%
 Suburban men 34% 36%
 Suburban women 31% 32%
 Rural men 32% 36%
 Rural women 29% 33%

Generally Speaking, Do You
Usually Think of Yourself as a ... 2000 (a) 2004 (b)

 Democrat

Total 33.7% 34.6%

Men 28.0% 28.9%
Women 38.8% 39.6%

High school or less 38% 38%
Some college 30% 31%
College degree or more 30% 33%
 Men high school or less 34% 33%
 Women high school or less 42% 42%
 Men some college 24% 26%
 Women some college 35% 36%
 Men college degree or more 23% 26%
 Women college degree or more 37% 40%

18-29 years old 31% 34%
30-44 years old 32% 31%
45-64 years old 35% 35%
65 and over 38% 39%
 Men 18-29 24% 27%
 Women 18-29 36% 40%
 Men 30-44 26% 26%
 Women 30-44 37% 37%

 Men 45-64 29% 30%
 Women 45-64 40% 40%
 Men 65 and over 33% 34%
 Women 65 and over 42% 42%

Household income below $35,000 40% 42%
$35,000 to less than $75,000 31% 33%
$75,000 and over 28% 29%
 Men below $35,000 35% 36%
 Women below $35,000 44% 45%
 Men $35,000 to $75,000 26% 28%
 Women $35,000 to $75,000 36% 38%
 Men $75,000 and over 23% 25%
 Women $75,000 and over 33% 35%

Whites 29% 30%
African Americans 68% 69%
Hispanic/Latinos 46% 44%
Asians 39% 37%
Native Americans 34% 36%
 White men 24% 25%
 White women 34% 34%
 African American men 62% 61%
 African American women 72% 75%
 Hispanic/Latino men 43% 39%
 Hispanic/Latino women 50% 49%

Union household 40% 44%
 Union men 35% 39%
 Union women 45% 49%
 Married 31% 31%
 Single 3% 41%
 Married men 27% 27%
 Married women 36% 36%
 Single men 31% 33%
 Single women 44% 46%

Northeast 36% 38%
Midwest 30% 33%
South 35% 34%
West 33% 33%
 Northeast men 31% 31%
 Northeast women 40% 44%
 Midwest men 25% 28%
 Midwest women 35% 38%
 Southern men 30% 29%
 Southern women 40% 39%
 Western men 26% 28%
 Western women 39% 38%

Urban 40% 42%
Suburban 31% 32%
Rural 32% 32%
 Urban men 33% 35%
 Urban women 46% 47%
 Suburban men 26% 26%
 Suburban women 36% 37%
 Rural men 28% 28%
 Rural women 37% 36%

 Independent
Generally Speaking, Do You
Usually Think of Yourself as a ... 2000 (a) 2004 (b)

Total 26.7% 25.2%

Men 30.0% 28.6%
Women 23.9% 22.2%

High school or less 25% 23%
Some college 27% 26%
College degree or more 29% 27%
 Men high school or less 29% 26%
 Women high school or less 22% 21%
 Men some college 31% 30%
 Women some college 25% 23%
 Men college degree or more 31% 31%
 Women college degree or more 26% 24%

18-29 years old 27% 27%
30-44 years old 26% 25%
45-64 years old 29% 26%
65 and over 24% 22%
 Men 18-29 30% 30%
 Women 18-29 24% 25%
 Men 30-44 28% 28%
 Women 30-44 24% 22%
 Men 45-64 32% 29%
 Women 45-64 25% 23%
 Men 65 and over 28% 26%
 Women 65 and over 21% 20%

Household income below $35,000 25% 24%
$35,000 to less than $75,000 28% 26%
$75,000 and over 28% 26%
 Men below $35,000 29% 28%
 Women below $35,000 23% 22%
 Men $35,000 to $75,000 30% 29%
 Women $35,000 to $75,000 25% 23%
 Men $75,000 and over 30% 29%
 Women $75,000 and over 25% 23%

Whites 28% 26%
African Americans 19% 17%
Hispanic/Latinos 21% 23%
Asians 28% 28%
Native Americans 31% 28%
 White men 31% 29%
 White women 25% 24%
 African American men 23% 24%
 African American women 15% 13%
 Hispanic/Latino men 24% 26%
 Hispanic/Latino women 19% 21%

Union household 27% 24%
 Union men 29% 27%
 Union women 24% 21%
 Married 27% 25%
 Single 27% 26%
 Married men 29% 28%
 Married women 24% 22%
 Single men 31% 30%
 Single women 23% 23%

Northeast 30% 28%
Midwest 29% 27%
South 25% 23%
West 24% 34%
 Northeast men 33% 31%
 Northeast women 27% 25%
 Midwest men 32% 31%
 Midwest women 26% 24%
 Southern men 28% 26%
 Southern women 21% 20%
 Western men 27% 28%
 Western women 22% 22%

Urban 26% 24%
Suburban 27% 26%
Rural 27% 25%
 Urban men 29% 28%
 Urban women 23% 21%
 Suburban men 30% 29%
 Suburban women 25% 23%
 Rural men 30% 27%
 Rural women 24% 23%

TABLE 5
Self-Identified Party by State-2004 versus 2000 among Registered Voters

 2004

State (N) Republican Democrat Independent (b)

Alabama (1,143) 35% 32% 26%
Arizona (1,140) 37% 30% 25%
Arkansas (675) 26% 37% 27%
California (6,225) 31% 37% 21%
Colorado (1,163) 33% 29% 31%
Connecticut (710) 22% 30% 41%
DC (150 (a)) NA NA NA
Delaware (205 (a)) NA NA NA
Florida (3,780) 35% 37% 23%
Georgia (1,871) 34% 32% 24%
Idaho (337) 38% 24% 28%
Illinois (2,579) 26% 37% 27%
Indiana (1,676) 36% 29% 27%
Iowa (894) 28% 30% 37%
Kansas (742) 43% 26% 26%
Kentucky (1,233) 37% 41% 16%
Louisiana (945) 32% 40% 24%
Maine (481) 30% 27% 37%
Maryland (1,217) 28% 44% 21%
Mass. (1,492) 16% 34% 47%
Michigan (2,486) 27% 35% 28%
Minnesota (1,535) 29% 31% 28%
Mississippi (571) 39% 36% 16%
Missouri (1,668) 29% 33% 26%
Montana (334) 38% 23% 30%
Nebraska (459) 45% 28% 23%
Nevada (419) 31% 32% 27%
New Hamp. (348) 25% 28% 44%
New Jersey (1,761) 27% 36% 30%
New Mexico (446) 31% 39% 25%
New York (4,105) 26% 41% 24%
N. Carolina (2,051) 35% 37% 21%
N. Dakota (184 (a)) NA NA NA
Ohio (3,250) 31% 35% 27%
Oklahoma (928) 37% 39% 18%
Oregon (1,047) 35% 33% 23%
Penn. (3,787) 36% 40% 18%
Rhode Is. (238 (a)) NA NA NA
S. Carolina (978) 36% 33% 21%
S. Dakota (217 (a)) NA NA NA
Tennessee (1,511) 32% 31% 26%
Texas (4,342) 37% 30% 23%
Utah (548) 44% 19% 26%
Vermont (218 (a)) NA NA NA
Virginia (1,840) 33% 29% 26%
Washington (1,666) 27% 34% 29%
W. Virginia (495) 33% 43% 16%
Wisconsin (1,539) 29% 34% 27%
Wyoming (148 (a)) NA NA NA

 2000

State (N) Republican Democrat Independent (b)

Alabama (855) 28% 35% 25%
Arizona (756) 40% 30% 23%
Arkansas (471) 19% 40% 32%
California (4,377) 32% 39% 19%
Colorado (737) 36% 25% 32%
Connecticut (527) 25% 29% 40%
DC (88 (a)) NA NA NA
Delaware (116 (a)) NA NA NA
Florida (2,433) 32% 35% 26%
Georgia (1,254) 30% 35% 24%
Idaho (284x) NA NA NA
Illinois (1,850) 25% 35% 28%
Indiana (1,001) 32% 28% 28%
Iowa (615) 29% 27% 35%
Kansas (584) 40% 26% 26%
Kentucky (721) 34% 39% 20%
Louisiana (863) 28% 46% 20%
Maine (333) 22% 27% 45%
Maryland (882) 25% 46% 22%
Mass. (984) 14% 32% 48%
Michigan (1,720) 27% 30% 29%
Minnesota (1,050) 25% 30% 34%
Mississippi (562) 32% 34% 20%
Missouri (1,094) 27% 29% 30%
Montana (177 (a)) NA NA NA
Nebraska (373) 44% 28% 24%
Nevada (278 (a)) NA NA NA
New Hamp. (242 (a)) NA NA NA
New Jersey (1,220) 26% 31% 34%
New Mexico (280 (a)) NA NA NA
New York (2,791) 27% 38% 25%
N. Carolina (1,493) 31% 38% 23%
N. Dakota (115 (a)) NA NA NA
Ohio (2,076) 32% 33% 27%
Oklahoma (677) 34% 44% 15%
Oregon (748) 29% 35% 26%
Penn. (2,230) 36% 37% 21%
Rhode Is. (184 (a)) NA NA NA
S. Carolina (791) 35% 28% 24%
S. Dakota (173 (a)) NA NA NA
Tennessee (1,061) 29% 34% 27%
Texas (3,098) 32% 31% 26%
Utah (333) 44% 17% 25%
Vermont (130 (a)) NA NA NA
Virginia (1,434) 29% 30% 30%
Washington (1,082) 25% 29% 34%
W. Virginia (389) 27% 48% 18%
Wisconsin (1,042) 26% 28% 32%
Wyoming (123 (a)) NA NA NA

 Difference

 2004-2000

State (N) Rep Dent Ind

Alabama (855) 7 -3 1
Arizona (756) -3 0 2
Arkansas (471) 7 -3 -5
California (4,377) -1 -2 2
Colorado (737) -3 4 -1
Connecticut (527) -3 1 1
DC (88 (a)) NA NA NA
Delaware (116 (a)) NA NA NA
Florida (2,433) 3 2 -3
Georgia (1,254) 4 -3 0
Idaho (284x) NA NA NA
Illinois (1,850) 1 2 -1
Indiana (1,001) 4 1 -1
Iowa (615) -1 3 2
Kansas (584) 3 0 0
Kentucky (721) 3 2 -4
Louisiana (863) 4 -6 4
Maine (333) 8 0 -8
Maryland (882) 3 -2 -1
Mass. (984) 2 2 -1
Michigan (1,720) 0 5 -1
Minnesota (1,050) 4 1 -6
Mississippi (562) 7 2 -4
Missouri (1,094) 2 4 -4
Montana (177 (a)) NA NA NA
Nebraska (373) 1 0 -1
Nevada (278 (a)) NA NA NA
New Hamp. (242 (a)) NA NA NA
New Jersey (1,220) 1 5 -4
New Mexico (280 (a)) NA NA NA
New York (2,791) -1 3 -1
N. Carolina (1,493) 4 -1 -2
N. Dakota (115 (a)) NA NA NA
Ohio (2,076) -1 2 0
Oklahoma (677) 3 -5 3
Oregon (748) 6 -2 -3
Penn. (2,230) 0 3 -3
Rhode Is. (184 (a)) NA NA NA
S. Carolina (791) 1 5 -3
S. Dakota (173 (a)) NA NA NA
Tennessee (1,061) 3 -3 -1
Texas (3,098) 5 -1 -3
Utah (333) 0 2 1
Vermont (130 (a)) NA NA NA
Virginia (1,434) 4 -1 -4
Washington (1,082) 2 5 -5
W. Virginia (389) 6 -5 -2
Wisconsin (1,042) 3 6 -5
Wyoming (123 (a)) NA NA NA

(a) Sample size too small for analysis.

(b) Independents are respondents who indicate they are "Independents."
This number excludes those who give a response of "something else,"
"don't know," or refused to answer.