GALLUP NEWS SERVICE
PRINCETON, NJ -- A special CNN/USA Today/Gallup 2002 election poll reveals that traditional voting disparities are still evident as this year's midterm elections approach. A gender gap is evident in the voting intentions of men and women, but much of this is due to the voting intentions of unmarried women, one of the most Democratic of subgroups in the U.S. population. Even larger differences are evident based on marital status, as both married men and married women are more likely to favor Republican candidates than are those who are not married. Currently, political independents are showing a preference for Democratic candidates in this fall's elections.
The poll, conducted Oct. 3-6, contains interviews with more than 1,500 adults nationwide, including 806 likely voters. Since the percentage of undecided and third-party voters varies by subgroup, the analysis here is based on only the two-party congressional vote among likely voters in each subgroup, providing an easier comparison across groups. This is not an exhaustive list of subgroups but does include those for which there is enough data to warrant analysis. Overall, the two-party vote among likely voters in this poll falls 50.5% in favor of Democratic candidates and 49.5% in favor of Republican candidates.
Key Groups
Partisanship is normally the strongest predictor of the American electorate's voting behavior. As such, it is important to determine how committed Republicans and Democrats are to their party's candidates, and how those without a party affiliation may vote. The data show that 94% of Republicans plan to vote for the Republican candidate in their district, and 96% of Democrats plan to vote for the Democratic candidate in their district, meaning both parties are doing a good job of holding their core supporters. Among independents, 61% say they will vote for the Democratic candidate and 39% for the Republican candidate, welcome news for the Democratic Party. Republicans, though, tend to vote at higher rates than Democrats and especially independents do, which means they are overrepresented in the pool of likely voters, and explains the closeness of the race despite independents' leanings. Similarly, 62% of those who describe themselves as ideological moderates say they would vote for the Democratic candidate.
In past elections, such as the 1996 presidential election, much has been made of a gender gap in American politics. That is, women tend to favor Democratic candidates while men are more likely to vote for Republicans. That gap is apparent in this year's congressional elections, too. Looking at the vote for major parties, 55% of female likely voters say they will vote for the Democratic candidate in their district while 45% will vote for the Republican. Among males, 53% of likely voters favor the Republican and 47% favor the Democrat. Thus, the gender gap is currently 8 points.
Two-Party Share of Congressional Vote Among Men and Women |
Oct. 3-6, 2002 | Likely voters |
In 2000, some political observers pointed out an emerging marriagegap in American politics. This is profoundly evident in the current data. Among those who are unmarried (including never married, divorced, and widowed), 61% prefer Democratic candidates and 39% prefer Republicans. But among those who are married, 55% support Republican congressional candidates and 45% support Democrats. The gap between the two groups is 16 points, twice as large as the gender gap. There is an interaction between gender and marital status as well: 68% of unmarried women plan to vote for Democratic candidates, while unmarried men are evenly divided in their vote intentions. Married men support Republican candidates by a 54% to 46% margin, and married women are even more Republican in their vote intentions, 58% to 42%. The gap between unmarried women and married women is 26 points.
Two-Party Share of Congressional Vote by Marital Status and Gender |
Oct. 3-6, 2002 | Likely voters |
Religion clearly has an effect on Americans' political behavior, and may explain part of the existence of the marriage gap (since those who are religious are more likely to get married and stay married than are those who are not religious). There are large differences in voting intentions between those who attend church regularly (that is, on a weekly basis, among whom 62% favor Republicans and 38% Democrats) and those who attend less regularly (among those who attend nearly weekly or monthly, or those who seldom or never attend, roughly six in 10 plan to vote for Democratic candidates).
Two-Party Share of Congressional Vote by Church Attendance |
Oct. 3-6, 2002 | Likely voters |
Protestants show a clear preference for Republican candidates, by a 58% to 42% margin. Catholics, a group that historically had been Democratic but has shown more competitive voting patterns in recent years, are among the more Democratic groups found in the poll, with 62% favoring Democratic candidates and 38% favoring Republicans. Preferences of practicing Catholics (those who attend church every week or nearly every week) are slightly less Democratic, 56% to 44%.
Two-Party Share of Congressional Vote by Religious Affiliation |
Oct. 3-6, 2002 | Likely voters |
As far as ageis concerned, voters between 18 and 29 show a clear preference for the Democratic Party. However, this group typically has very low levels of turnout, and therefore, very few people in this age group are deemed to be likely to vote. Senior citizens (aged 65 and older) also show a preference for the Democratic Party, but among likely voters in this group, the vote is fairly evenly divided between Republican and Democratic candidates. However, those between the ages of 50 and 64 (likely voters and otherwise) show a clear preference for Democratic candidates (56% to 44%). The group of Americans between the ages of 30 and 49, the largest group of voters in this demographic category, prefer Republican candidates by a 54% to 46% margin.
As is usually the case, voters differ depending on their place of residence. Fifty-eight percent of voters in urban areas say they will vote for the Democratic candidate, while 42% say they will vote for the Republican. The reverse is true in rural areas, where 56% favor Republicans and 44% Democrats. Suburban voters -- the largest bloc of voters in this category -- are essentially equally divided in their vote, with 51% favoring Democrats and 49% Republicans. The close split among suburban residents hides evidence of a gender gap -- 55% of suburban women favor Democrats while 55% of suburban men favor Republicans.
Strongest Democratic Groups
The following table shows the groups that most strongly support Democratic candidates for Congress. Notable is the overwhelming support of black voters, who nearly unanimously say they intend to vote for Democrats this fall. Members of low-income households and Americans with postgraduate educations also rank among the most Democratic groups.
Strongest Supporters of Democratic Congressional Candidates Among Likely Voters
Group |
Democratic Share of Two-Party Vote |
Republican Share of Two-Party Vote |
% |
% |
|
Blacks |
96 |
4 |
Democrats |
96 |
4 |
Liberals |
91 |
9 |
Nonwhites |
76 |
24 |
Unmarried women |
68 |
32 |
Urban women |
64 |
36 |
Moderates |
62 |
38 |
Females with no college education |
62 |
38 |
Postgraduate education |
62 |
38 |
Catholics |
62 |
38 |
Unmarried |
61 |
39 |
Less than $30,000 household income |
61 |
39 |
Strongest Republican Groups
The single strongest group in supporting Republican candidates is, not surprisingly, Republican Party identifiers. While those with a postgraduate education are among the strongest Democratic groups, those with a college degree but no postgraduate work and those who attended college but do not have a degree are among the most Republican groups. Married people with children are even more likely to say they will support Republican candidates than are the larger group of married people.
Strongest Supporters of Republican Congressional Candidates Among Likely Voters
Group |
Republican Share of Two-Party Vote |
Democratic Share of Two-Party Vote |
% |
% |
|
Republicans |
94 |
6 |
Conservatives |
73 |
27 |
Rural women |
64 |
36 |
Weekly church attendees |
62 |
38 |
College graduates (no postgrad. education) |
59 |
41 |
Practicing Protestants |
59 |
41 |
Married with children under 18 |
59 |
41 |
Protestants |
58 |
42 |
Married women |
58 |
42 |
Household income $50,000-$74,999 |
57 |
43 |
Two-income family with children under 18 |
57 |
43 |
Some college education (no degree) |
57 |
43 |
Below is the full rundown of all groups analyzed.
Survey Methods
These results are based on telephone interviews with a randomly selected national sample of 1,502 national adults and 806 likely voters, 18 years and older, conducted Oct. 3-6, 2002. For results based on the national adult sample, one can say with 95 percent confidence that the maximum error attributable to sampling and other random effects is plus or minus 3 percentage points. For results based on the likely voter sample, one can say with 95 percent confidence that the maximum error attributable to sampling and other random effects is plus or minus 4 percentage points. In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.
Group |
Two-Party Vote for Congress |
|
% Democrat |
% Republican |
|
Black |
95.9 |
4.1 |
Democrat |
95.9 |
4.1 |
Liberal |
90.5 |
9.5 |
Nonwhite |
76.3 |
23.7 |
Unmarried women |
67.7 |
32.3 |
Urban women |
63.6 |
36.4 |
Moderate |
62.1 |
37.9 |
Female, no college |
62.1 |
37.9 |
Postgrad. |
61.5 |
38.5 |
Catholic |
61.5 |
38.5 |
Unmarried |
61.3 |
38.7 |
Less than $30,000 |
60.8 |
39.2 |
Independent |
60.7 |
39.3 |
Unmarried, no children |
60.6 |
39.4 |
Attend church monthly |
60.4 |
39.6 |
Union member |
58.9 |
41.1 |
Urban |
57.7 |
42.3 |
Seldom/Never attend church |
57.3 |
42.7 |
East |
56.8 |
43.2 |
HS or less |
56.7 |
43.3 |
Midwest |
56.5 |
43.5 |
Female college grad |
56.4 |
43.6 |
Practicing Catholic |
55.8 |
44.2 |
Do not own stock |
55.8 |
44.2 |
50-64 |
55.7 |
44.3 |
Suburban women |
55.2 |
44.8 |
Swing voter |
55.0 |
45.0 |
Women |
54.7 |
45.3 |
18-29 |
54.6 |
45.4 |
No children |
53.7 |
46.3 |
Female, some college |
52.2 |
47.8 |
Not employed |
52.1 |
47.9 |
Working moms |
51.6 |
48.4 |
Committed voter |
51.0 |
49.0 |
$30,000-$49,999 |
50.5 |
49.5 |
National adults |
50.5 |
49.5 |
Employed |
50.5 |
49.5 |
Suburban |
50.5 |
49.5 |
Male college grad |
50.0 |
50.0 |
Male, no college |
50.0 |
50.0 |
Unmarried men |
50.0 |
50.0 |
Two incomes, w/o kids |
50.0 |
50.0 |
Urban men |
50.0 |
50.0 |
$75,000 and above |
49.5 |
50.5 |
65+ |
49.5 |
50.5 |
Retired |
49.0 |
51.0 |
Non-union member |
49.0 |
51.0 |
Stock owner |
49.0 |
51.0 |
Rural men |
48.9 |
51.1 |
One-income family |
48.0 |
52.0 |
Men |
47.4 |
52.6 |
Have children under 18 |
47.4 |
52.6 |
South |
46.9 |
53.1 |
Two-income family |
46.4 |
53.6 |
30-49 |
46.3 |
53.7 |
Married men |
46.3 |
53.7 |
Male, some college |
45.7 |
54.3 |
West |
45.3 |
54.7 |
Suburban men |
44.9 |
55.1 |
Married |
44.8 |
55.2 |
White |
44.7 |
55.3 |
Member of Religious Right |
44.3 |
55.7 |
Rural |
43.8 |
56.2 |
Some college |
43.2 |
56.8 |
Two incomes, have children under 18 |
43.2 |
56.8 |
$50,000-$74,999 |
42.7 |
57.3 |
Married women |
42.3 |
57.7 |
Protestant |
42.3 |
57.7 |
Married with children under 18 |
41.2 |
58.8 |
Practicing Protestant |
41.1 |
58.9 |
College grad only |
40.9 |
59.1 |
Attend church weekly |
38.5 |
61.5 |
Rural women |
36.0 |
64.0 |
Conservative |
26.8 |
73.2 |
Republican |
6.1 |
93.9 |