By early 2004, party ID showed some signs of returning to "normal," with the Democrats back in the lead. As shown in Table 6, however, the Democrats were not successful in improving their position across the board demographically.
Table 6: The
Democratic Party ID Rebound (Early 2004) By Demographics |
|
2001-2002
-Early Post 9/11- |
2004 Dem Primaries/
Pre-Dem Convention |
|
|
|
Rep |
Dem |
Rep |
Dem |
Margin Shift |
Minimum N |
|
% |
% |
% |
% |
%/party |
|
Total |
32 |
31 |
30 |
33 |
+4D |
6,460 |
|
|
|
|
|
|
|
Men |
33 |
27 |
32 |
29 |
+3D |
3,124 |
Women |
31 |
35 |
28 |
37 |
+5D |
3,336 |
|
|
|
|
|
|
|
White |
37 |
26 |
35 |
28 |
+4D |
5,204 |
Black |
10 |
64 |
7 |
64 |
+3D |
518 |
Hispanic |
25 |
36 |
26 |
34 |
+3R |
372 |
|
|
|
|
|
|
|
18-29 |
28 |
28 |
26 |
30 |
+4D |
1,073 |
30-49 |
34 |
30 |
31 |
33 |
+6D |
2,185 |
50-64 |
33 |
31 |
31 |
34 |
+5D |
1,745 |
65+ |
32 |
40 |
33 |
38 |
+3R |
1,272 |
|
|
|
|
|
|
|
College grad+ |
37 |
29 |
33 |
33 |
+8D |
2,447 |
Some college |
33 |
28 |
32 |
32 |
+5D |
1,568 |
HS or less |
29 |
34 |
28 |
35 |
+2D |
2,352 |
|
|
|
|
|
|
|
Northeast |
29 |
33 |
27 |
35 |
+4D |
1,148 |
Midwest |
32 |
28 |
29 |
33 |
+8D |
1,560 |
South |
32 |
34 |
32 |
35 |
+1D |
2,313 |
West |
33 |
28 |
31 |
29 |
+3D |
1,439 |
|
|
|
|
|
|
|
Southern white |
39 |
26 |
39 |
27 |
+1D |
1,768 |
Non-South white |
35 |
26 |
33 |
29 |
+5D |
3,436 |
|
|
|
|
|
|
|
Urban |
26 |
37 |
24 |
38 |
+3D |
1,853 |
Suburban |
34 |
28 |
32 |
32 |
+6D |
3,215 |
Rural |
34 |
29 |
33 |
30 |
+2D |
1,392 |
Most subgroups shifted toward the Democrats by a two to six point margin, but some did not move toward the party at all. Hispanics and seniors ages sixty-five and over actually had a more Republican distribution in early 2004 polling than they did in polling from the early post-9/11 period used for comparisons. Southern whites, whose party ID basically stayed the same after 9/11, also didn't move in early 2004--perhaps an early indicator that John Edwards's addition to the Democratic ticket would do nothing to prevent a sweep of the South by the Republicans in November.
The final major shift in party ID selected for demographic analysis is the post-GOP convention surge. As seen in Table 7, the demographic patterns were different in a number of important ways from the patterns observed in the post-9/11 GOP surge.
Table 7: The Post-GOP Convention Surge (Sept. 2004) By Demographics |
|
2004 Dem Primaries/
Pre-Dem Convention |
2004 Post-GOP
--Convention-- |
|
|
|
Rep |
Dem |
Rep |
Dem |
Margin Shift |
Minimum N |
|
% |
% |
% |
% |
%/party |
|
Total |
30 |
33 |
33 |
31 |
+5R |
2,354 |
|
|
|
|
|
|
|
Men |
32 |
29 |
32 |
27 |
+2R |
1,135 |
Women |
28 |
37 |
34 |
34 |
+9R |
1,219 |
|
|
|
|
|
|
|
White |
35 |
28 |
38 |
25 |
+6R |
1,932 |
Black |
7 |
64 |
- |
- |
- |
156 |
Hispanic |
26 |
34 |
- |
- |
- |
130 |
|
|
|
|
|
|
|
18-29 |
26 |
30 |
27 |
28 |
+3R |
313 |
30-49 |
31 |
33 |
37 |
29 |
+10R |
800 |
50-64 |
31 |
34 |
33 |
32 |
+4R |
654 |
65+ |
33 |
38 |
33 |
36 |
+2R |
522 |
|
|
|
|
|
|
|
College grad+ |
33 |
33 |
38 |
29 |
+9R |
928 |
Some college |
32 |
32 |
36 |
30 |
+6R |
568 |
HS or less |
28 |
35 |
30 |
32 |
+5R |
825 |
|
|
|
|
|
|
|
Northeast |
27 |
35 |
27 |
31 |
+4R |
461 |
Midwest |
29 |
33 |
33 |
29 |
+8R |
562 |
South |
32 |
35 |
38 |
30 |
+11R |
824 |
West |
31 |
29 |
31 |
33 |
+4R |
507 |
|
|
|
|
|
|
|
Southern white |
39 |
27 |
46 |
23 |
+11R |
645 |
Non-South white |
33 |
29 |
35 |
26 |
+5R |
1,287 |
|
|
|
|
|
|
|
Urban |
24 |
38 |
30 |
34 |
+10R |
615 |
Suburban |
32 |
32 |
35 |
29 |
+6R |
1,158 |
Rural |
33 |
30 |
34 |
30 |
+1R |
581 |
Note: Too few cases to report for black and hispanic subgroups
While white southerners had not moved their party ID in the post-9/11 period up to this point, after the GOP convention they shifted toward the Republicans by a significant margin (eleven percentage points). The speech at the GOP convention by southern Democrat Zell Miller critical of his party was criticized by many in the media as too harsh. But Miller's speech, and the tough speech by Dick Cheney as well, appear to have resonated with white southerners.
Other subgroups that stand out as moving toward the Republicans at this point in election year 2004 include urban residents (+10 percentage points), those ages thirty to forty-nine (+10), women (+9), and college graduates (+9). The improvement for the Republicans (and Bush's standing in the horserace polls) among women and those ages thirty to forty-nine may have contributed to the notion that "security moms"--for whom concerns about terrorism trumped considerations like health care, education, and other social issues--were key to the election. Our analysis, however, shows that the movement of thirty to forty-nine-year-olds toward the Republicans after the GOP convention was no greater among women in this age group than it was among men. It was not "security moms," specifically, but women overall, whose greater support for Bush and the Republicans in this election year was a key to the outcome. The national exit poll showed Bush behind Kerry by just three points among women voters (51 percent to 48 percent), while in 2000 he lost the women's vote to Al Gore by eleven points (fifty-four percent to forty-three percent).
Next issue of Public Opinion Pros: Historical precedents to party ID movement around conventions, and the horserace.
Larry Hugick is chairman and Stacy DiAngelo is assistant study director of Princeton Survey Research Associates International. |