### Who Voted for the Dems?

So, just who provided the Democrats with their historic victory yesterday?

One way to get a handle on this issue is to compare the house exit polls from the 2006 and 2004 elections.

To demonstrate how this can be done, consider the gender gap:

Vote by Gender

2006    D    R
Male (49%)    50%    47%
Female (51%)    55%    43%

2004    D    R
Male (46%)    45%    53%
Female (54%)     52%    46%

In 2004, 46% of voters were men and they broke 45-53 for the Republicans. In 2006, 49% of voters were men and they broke 50-47 for the Democrats.

So how much better did the Democrats do with men in '06 compared to '04? One way to measure this is to estimate how much of the swing in the total vote to the Democrats is accounted for by these male voters compared to women.

In 2004 the Democrats' share of the House vote was 48.8%, in 2006 it was 52.6%, an increase of 3.8 percentage points. How much of that 3.8 percentage point increase is accounted for by men? It would be

%total vote '06 x %Dem '06 - %total vote '04 x %Dem '04 =
(.49) x (.50) - (.46) x (.45) = 3.8

So men accounted for all of the 3.8 point swing to the Dems and women accounted for none of it. This is due not only to the fact that the Democrats increased their share of the male vote by 5 points between 2004 and 2006, but also that the male share of total vote increased from 46% to 49%. So this approach can account for turnout effects as well.

As for women, the Democrats did improve from 52% to 55%, but the female share of the total vote fell from 54% to 51%, so the (reduced) relative turnout effect counteracted the (pro-Dem) change in voting preference in this case.

What happens when we look at various voting categories this way?

The Democrats did very well among men:

VOTE BY GENDER
Male (49%)     3.8%
Female (51%)     0.0%

And white men:

VOTE BY RACE AND GENDER
White Men (39%)     3.5%
White Women (40%)     1.2%
Non-White Men (9%)     -0.1%
Non-White Women (11%)     -1.3%

Overall, improvement among white voters was key, but so was the overwhelming endorsement that Latino voters gave to the Democrats (the Dems' share of the Latino vote rose from 55% in '04 to 69% in '06). How's that immigration issue working for the Rethugs?

VOTE BY RACE
White (79%)     4.8%
Latino (8%)     1.1%
Asian (2%)     0.1%
Other (2%)     0.0%
African-American (10%)     -1.8%

In terms of age, perhaps surprisingly, voters 45 and older accounted for more than their share of the swing to the Dems. Both a turnout effect and a preference effect is working here. How's that social security privatization issue working for the Rethugs?

VOTE BY AGE
45-59 (34%)     3.3%
60 and Older (29%)     3.0%
30-44 (24%)     -0.9%
18-29 (12%)     -1.6%

The Democrats saw noticable improvement with middle class and upper-middle class voters:

VOTE BY INCOME
\$75-100,000 (16%)     2.0%
\$100-150,000 (13%)     1.4%
\$50-75,000 (22%)     0.9%
\$200,000 or More (5%)     0.8%
\$150-200,000 (5%)     0.7%
\$30-50,000 (21%)     0.5%
Under \$15,000 (7%)     -0.4%
\$15-30,000 (12%)     -1.4%

And also with more highly-educated voters:

VOTE BY EDUCATION
Some College (31%)     0.8%
No High School (3%)     0.0%

Democrats did better among non-union voters:

ARE YOU A UNION MEMBER?
No (86%)     3.4%
Yes (14%)     0.4%

ANYONE IN HOUSEHOLD IN A UNION?
No (77%)     4.3%
Yes (23%)     0.1%

In general, the economy does not seem to have been so important an issue for the Democrats. In fact, most of their gains were made among people who thought both the economy and their personal financial situation were doing well:

NATIONAL ECONOMY
Good (40%)     5.5%
Not Good (37%)     2.5%
Excellent (9%)     0.6%
Poor (13%)     -4.1%

FAMILY'S FINANCIAL SITUATION
Same (44%)     5.1%
Better (30%)     1.4%
Worse (25%)     -2.3%

Improvements with independents and moderates played a key role in the Democratic victory:

VOTE BY PARTY ID
Independent (26%)     2.6%
Democrat (38%)     1.1%
Republican (36%)     0.2%

VOTE BY IDEOLOGY
Moderate (47%)     3.0%
Conservative (32%)     0.6%
Liberal (20%)     0.0%

As for religion, secular voters provided more of the swing to Democrats than any other group, followed closely by Protestants and Catholics, although the Democrats also did better with those who attend church weekly:

VOTE BY RELIGION
None (11%)     1.6%
Protestant (55%)     1.5%
Catholic (26%)     1.1%
Jewish (2%)     -0.5%
Other (6%)     -0.6%

WHITE EVANGELICAL/BORN-AGAIN?  &nbsp ;
No (76%)     2.5%
Yes (24%)     1.0%

VOTE BY CHURCH ATTENDANCE
Weekly (28%)     2.0%
Never (15%)     1.1%
More Than Weekly (17%)     0.5%
Monthly (12%)     0.3%
A Few Times a Year (25%)     -0.4%

The Democrats succeeded in narrowing the so-called "marriage gap" somewhat, although most gains were made among married couples without children:

ARE YOU MARRIED?
Yes (68%)     5.6%
No (32%)     -1.4%

ARE YOU MARRIED WITH CHILDREN?
No (73%)     4.2%
Yes (27%)     1.2%

DO YOU HAVE CHILDREN UNDER 18?
No (66%)     4.3%
Yes (34%)     0.0%

Opinions about Bush (and the direction of the country as a whole) mattered quite a bit, not surprisingly:

HOW GEORGE W. BUSH IS HANDLING HIS JOB
Strongly Disapprove (41%)     5.5%
Somewhat Disapprove (15%)     0.7%
Somewhat Approve (23%)     -1.1%
Strongly Approve (19%)     -1.3%

OPINION OF BUSH
Angry (29%)     6.0%
Dissatisfied (30%)     0.2%
Enthusiastic (12%)     -0.3%
Satisfied (27%)     -1.7%

IS U.S. GOING IN RIGHT DIRECTION?
No (55%)     4.8%
Yes (41%)     -0.3%

Opinions about the Iraq War mattered greatly. Were there actually people who thought the Dems should ditch the war as an issue and run on the economy?

U.S. WAR IN IRAQ
Strongly Disapprove (39%)     5.8%
Somewhat Disapprove (16%)     -0.6%
Strongly Approve (19%)     -0.6%
Somewhat Approve (23%)     -1.8%

DID WAR IN IRAQ IMPROVE U.S. SECURITY?
No (59%)     5.9%
Yes (35%)     -1.7%

This is an interesting one -- late deciders broke for the Democrats, as in '04, but there were many more late deciders this time than last time. In 2004, fully 80% of voters had already made up their minds who to vote for more than a month before the election, compared to only 50% this time. In 2006, 28% made up their minds in the last week compared to only 10% on '04:

WHEN DID YOU DECIDE YOUR HOUSE VOTE?
Last Month (21%)     5.8%
Last Week (9%)     3.5%
Today (10%)     3.2%
Last Three Days (9%)     3.1%
Before Then (50%)     -10.6%

What about where voters lived? Democrats did well among suburban, small-city, and rural voters:

VOTE BY SIZE OF COMMUNITY
Suburbs (47%)     2.4%
Smaller Cities (20%)     1.9%
Rural (18%)     1.8%
Big Cities (10%)     -0.6%
Small Towns (5%)     -1.3%

They also did well everywhere but the South:

VOTE BY REGION
West (21%)     1.7%
Northeast (22%)     1.3%
Midwest (27%)     1.3%
South (30%)     -0.3%

Tags: Election 2006 (all tags)

1 Comment

##### Re: Who Voted for the Dems?
fascinating
Great report.
Thanks
by djjimz 2006-11-09 02:37AM | 0 recs