skip to main content
Obama's Appeal to Well-Educated Not Conducive to Winning Nomination

Obama's Appeal to Well-Educated Not Conducive to Winning Nomination

Dukakis only candidate who appealed more to well-educated who has won since 1988

GALLUP NEWS SERVICE

PRINCETON, NJ -- Illinois Sen. Barack Obama is a clear second place behind New York Sen. Hillary Clinton in the race for the Democratic presidential nomination, but he is highly competitive with Clinton among the most educated segment of the party. That appeal may be one reason he has met or surpassed Clinton's fundraising totals despite not gaining much ground in voter support this year -- well-educated Americans tend to have greater income.

An analysis of historical Gallup Poll data on rank-and-file Democrats' nomination preferences shows that at least one candidate has exhibited a pattern similar to Obama's education skew in each election cycle since 1988, but that candidate usually does not end up winning the Democratic presidential nomination.

Obama's Support

According to combined data from Gallup's national Democratic nomination trial heat polls conducted in July and August 2007, Obama's support rises from 19% among Democrats with a high school education or less, to 28% of those who attended college but did not finish, and 33% among college graduates.  

By contrast, Clinton's support shows a downward trend by education level, as 51% of Democrats with a high school education or less, 45% of those with some college education, and 33% of college graduates support her. Thus, while Clinton leads Obama by 32 percentage points (51% to 19%) among Democrats with the least formal education, she merely ties him among the most educated Democrats.

Former North Carolina Sen. John Edwards' support is a consistent 14% among Democrats of all education levels.

 

Analyses of historical Gallup Poll data leading up to the Iowa caucuses show Obama's support pattern by education is not uncommon. The analyses rely on combined results of available data from national polls conducted in the November, December, and January months prior to the Iowa caucus, usually held in mid-to-late January of the presidential election year. Thus, the figures show where the nomination contest race stood among all Democrats before actual voting took place.

In some cases, such as in 1988, 1992, and 2004, the pre-Iowa figures were not predictive of the eventual outcome, as the candidate leading the national trial heat polls did not win the nomination. In 1996, President Bill Clinton was unchallenged for the nomination so there is no data for that election.

Also, to ensure comparability across elections and maintain robust sample sizes, the analysis compares the preferences of Democrats without any college education and those who attended college. In the current campaign, Hillary Clinton's support is 51% among Democrats without any college education and 39% among those who attended college. Obama's support is 19% among non-college educated Democrats and 31% among Democrats who attended college.

2004

By the fall of 2003, former Vermont Gov. Howard Dean was able to translate his fundraising prowess -- fueled in large part by Internet donations -- into front-runner status for the 2004 Democratic nomination. Prior to the Iowa caucuses, 23% of Democrats nationwide named Dean as their top choice for the party's presidential nomination, followed by Retired Gen. Wesley Clark at 16%, Connecticut Sen. Joe Lieberman at 12%, and eventual nominee Massachusetts Sen. John Kerry tied with Missouri Rep. Dick Gephardt at 9%.

Dean's support showed a decided education skew -- 16% of non-college educated Democrats and 27% of Democrats with college educations (including 33% of college graduates) supported him. Although Kerry's pre-Iowa national support was still limited, college educated (9%) and non-college educated Democrats (9%) were equally likely to support him.

Support for 2004 Democratic Presidential Nomination
by Education Level
(Nov 2003-Jan 2004 Gallup Polls)

Candidate

 

All Democrats

Democrats with no college education
(percentage
of support)

Democrats with college education
(percentage
of support)

%

%

%

Howard Dean

23

16

27

Wesley Clark

16

14

17

Joe Lieberman

12

11

12

Dick Gephardt

9

11

9

John Kerry

9

9

9

John Edwards

7

8

5

Kerry surprised everyone by winning the Iowa caucuses while Dean finished a disappointing third, and Dean didn't help his cause by punctuating his concession speech with the infamous "Dean scream." Kerry followed up the Iowa win by defeating Dean in New Hampshire, seriously crippling the former front-runner's candidacy.

2000

Vice President Al Gore was the clear front-runner in 2000, and his only notable challenger was former New Jersey Sen. Bill Bradley. Gore had a significant lead over Bradley in pre-Iowa national trial heat polls, including among all education groups. Despite the considerable deficit, Bradley did get greater support among college-educated Democrats (36%) than among non-college educated Democrats (29%). Bradley's support stretched to as high as 41% among Democrats with post-graduate education, but he still trailed Gore (49%) among this group.

Support for 2000 Democratic Presidential Nomination
by Education Level
(Nov 1999-Jan 2000 Gallup Polls)

Candidate

 

All Democrats

Democrats with no college education
(percentage
of support)

Democrats with college education
(percentage
of support)

%

%

%

Al Gore

57

60

54

Bill Bradley

33

29

36

After a decisive victory for Gore in the Iowa caucuses, he edged out Bradley in New Hampshire, one of the few states in which experts thought Bradley could defeat Gore. Gore wound up winning every Democratic primary and caucus that year.

1992

The 1992 Democratic field was largely filled with unknown candidates after better-known candidates such as former New York Gov. Mario Cuomo, Bradley, and Gore declined to run. Indeed, no front-runner had emerged before Iowa and New Hampshire. Former California Gov. Jerry Brown, out of office for nearly 10 years and who had twice unsuccessfully sought the presidency, was the best known and had a slight edge over Bill Clinton in rank-and-file Democrats' preferences as 1992 began.

Some believed that former Vietnam veteran and then Nebraska Sen. Bob Kerrey might emerge from the field. Kerrey showed the greatest education skew among the candidates, with his support nearly twice as high among Democrats who attended college (15%) as those with a high school education or less (8%).

Support for 1992 Democratic Presidential Nomination
by Education Level
(Jan 3-9, 1992, Gallup Poll)

Candidate

 

All Democratic
registered voters

Democratic
registered voters
with no college education
(percentage
of support)

Democratic
registered voters
with college
education
(percentage
of support)

%

%

%

Jerry Brown

21

21

22

Bill Clinton

17

16

19

Bob Kerrey

11

8

15

Doug Wilder

9

11

7

Tom Harkin

9

9

8

Paul Tsongas

6

6

6

Kerrey finished a distant third behind Paul Tsongas and Clinton in New Hampshire, but went on to win the South Dakota caucuses. He performed poorly in the remaining primaries and caucuses before ending his campaign in early March.

1988

In 1988, two Democratic candidates showed greater appeal to college-educated than non-college educated party members. One was Illinois Sen. Paul Simon; the other was the eventual nominee, Massachusetts Gov. Michael Dukakis. Dukakis' support was twice as high among college-educated Democrats nationally (15%) as among non-college educated Democrats (7%). Simon's appeal was three times greater among college-educated Democrats (13%) than non-college educated Democrats (5%). (Bruce Babbitt's support also showed an education skew, but only 2% of Democrats supported him, so he is not included in this analysis.)

Support for 1988 Democratic Presidential Nomination
by Education Level
(Dec 16-17, 1987, and Jan 8-17, 1988, Gallup Polls)

Candidate

 

All Democrats

Democrats with no college education
(percentage
of support)

Democrats with college education
(percentage
of support)

%

%

%

Gary Hart

26

30

20

Jesse Jackson

18

18

17

Michael Dukakis

10

7

15

Paul Simon

8

 5

13

Al Gore

4

4

4

Dick Gephardt

4

4

4

Gary Hart had left the campaign due to accusations of marital infidelity in May 1987 but renewed his presidential bid in December and immediately regained his front-runner status nationally. His national support did not foreshadow his dismal performance in New Hampshire, where he finished seventh with Dukakis winning that primary. A spirited campaign saw most candidates win at least one state's primary or caucus during the process before Dukakis emerged as the nominee.

Implications

Obama and future candidates who have a stronger appeal to educated Democrats can look to Dukakis' campaign as a model for success, and attempt to avoid the more common fate among past candidates whose support increased by respondent education level.   

It is not entirely clear why the candidates whose support is positively correlated with education have not fared better in recent campaigns. But at the most basic level, these candidates were typically not that well positioned at the start of the primary and caucus season, and thus started out at a disadvantage. Dean is an obvious exception to that, but Bradley and Kerrey, for example, were well behind the front-runners and were never able to gain the momentum needed to catch or overtake them.   

This may be due in large part to name recognition. Especially early in the campaign, Democrats with a college education are probably more likely than Democrats without a college education to be familiar with the lesser-known candidates. Name identification is a critical factor in candidate support because respondents are not likely to support a candidate they know little about. Much of the battle for the candidates is to become known, and once they do, the campaign can play out differently from how it starts.

At this point in the 2008 campaign, Clinton is nearly universally known among all Democrats, regardless of education level. On the other hand, Obama is a much more familiar to highly educated Democrats than those with less formal education.

Alternatively, the candidates who have shown growing support across education levels may share similarities -- such as in their issue positions or personal style -- that appeal to highly educated Democrats but may not attract the broader base of the party.

Additionally, the candidates who exhibit an education skew in their support may just have run bad campaigns -- Bradley, Dean, Kerrey, and Simon were all thought to have run lackluster campaigns. Dukakis is the obvious exception to this.

It is also important to note that the candidate with greater appeal to educated Democrats often is not even the top choice among this group. While college-educated Democrats were more likely to support Bradley, Kerrey, Dukakis, and Simon than non-college educated Democrats, college-educated Democrats still preferred Gore to Bradley, Brown or Bill Clinton to Kerrey, and Hart or Jackson to Simon and Dukakis at the outset of the campaign.

While Obama's pattern of support certainly does not mean his candidacy is doomed, he would be bucking the recent historical trend of Democratic nomination outcomes should he win.

Survey Methods

Results for 2007 are based on telephone interviews with a randomly selected national sample of 1,489 Democrats and Democratic leaners, aged 18 and older, conducted in July and August 2007. This sample includes 414 Democrats who did not attend college and 1,060 who attended college. For results based on the total sample, one can say with 95% confidence that the maximum error attributable to sampling and other random effects is ±3 percentage points.

Results for 2004 are based on telephone interviews with a randomly selected national sample of 3,147 Democrats and Democratic leaners, aged 18 and older, conducted from November 2003 through January 11, 2004. This sample includes 983 Democrats who did not attend college and 2,151 who attended college. For results based on the total sample, one can say with 95% confidence that the maximum error attributable to sampling and other random effects is ±3 percentage points.

Results for 2000 are based on telephone interviews with a randomly selected national sample of 3,190 Democrats and Democratic leaners, aged 18 and older, conducted from November 1999- through January 19, 2000. This sample includes 1,128 Democrats who did not attend college and 2,046 who attended college. For results based on the total sample, one can say with 95% confidence that the maximum error attributable to sampling and other random effects is ±3 percentage points.

Results for 1992 are based on telephone interviews with a randomly selected national sample of 1,214 Democrats and Democratic leaners who were registered to vote, aged 18 and older, conducted from January 3-9, 1992. This sample includes 576 Democrats who did not attend college and 635 who attended college. For results based on the total sample, one can say with 95% confidence that the maximum error attributable to sampling and other random effects is ±3 percentage points.

Results for 1988 are based on telephone interviews with a randomly selected national sample of 1,423 Democrats and Democratic leaners, aged 18 and older, conducted December 16-17, 1987 and January 8-17, 1988. This sample includes 781 Democrats who did not attend college and 638 who attended college. For results based on the total sample, one can say with 95% confidence that the maximum error attributable to sampling and other random effects is ±3 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.


Gallup https://news.gallup.com/poll/28414/Obamas-Appeal-WellEducated-Conducive-Winning-Nomination.aspx
Gallup World Headquarters, 901 F Street, Washington, D.C., 20001, U.S.A
+1 202.715.3030