PRINCETON, NJ -- On Tuesday, Massachusetts Sen. John Kerry did what almost everyone expected, and what the CNN/USA Today/Gallup pre-election poll had estimated would be the case: he won the New Hampshire primary by a substantial margin over former Vermont Gov. Howard Dean, capping a rise to victory in the Granite State that was fueled by his surprising victory in Iowa a week earlier.
The close race for third among retired Gen. Wesley Clark, and Sens. John Edwards (N.C.) and Joe Lieberman (Conn.), also transpired as predicted by pre-election polls. The final figures showed less than a four-point spread between Clark, who took third, and Lieberman, who came in fifth.
There was a great deal of discussion before the New Hampshire primary about the historically volatile New Hampshire voters, with assumptions that they would be rethinking their votes and changing their minds even as they walked into the voting booths. But that didn't happen. Voters' preferences indeed did undergo significant change in the days after the Iowa caucuses, as Kerry moved up rapidly in the polls and as Dean (and Clark) fell. But by Saturday, three days before the primary election, the race stayed remarkably stable until Election Day. That was reflected in the closeness of Gallup's final estimate -- based on Saturday and Sunday interviewing -- to the final vote outcome.
The final results showed Kerry with 38% of the vote, Dean with 26%, Clark with 12%, Edwards with 12%, and Lieberman with 9%. Sunday's final estimate from the CNN/USA Today/Gallup poll was Kerry 36%, Dean 25%, Clark 13%, Edwards 10%, and Lieberman 10%, with 4% undecided.
Why Kerry Won
Both the CNN/USA Today/Gallup pre-election poll and the results of the network-sponsored exit poll conducted by Edison Media Research and Mitofsky International suggest that Kerry won because many voters were looking for a candidate who had the best chance of beating President Bush in the November election, and perceived Kerry to be that candidate.
The CNN/USA Today/Gallup tracking poll, for example, showed that over half of the New Hampshire voters, 56%, perceived Kerry as having the best chance of beating Bush. Only 16% of the voters felt that about Dean, and smaller numbers felt that about the other candidates. In other words, whatever else they may have thought about Kerry's strengths, New Hampshire voters became convinced that he is the one who is most electable in November. And that seems to be a very important issue to Democratic voters this year. The exit polls sponsored by the media after Tuesday night's primary confirm this driving factor in the primary. Of those voters who said that the ability to defeat George W. Bush was the most important candidate characteristic in their vote, 62% voted for Kerry, compared to only 10% who voted for Dean.
Dean apparently still retained a considerable amount of good will among New Hampshire Democratic primary voters even as he lost to Kerry. The weekend CNN/USA Today/Gallup poll contained a series of questions that asked voters which of the candidates was best described by a series of five positive personal characteristics. Dean was more likely than Kerry to be selected on three of the dimensions, by three- to five-point margins, and trailed Kerry by only two points on a fourth. These include such things as "has new ideas that would help solve the country's problems," "shares your values," and "is in touch with the average American."
Indeed, the exit poll showed that Dean was the favored candidate among New Hampshire primary voters who said that standing up for what one believes in or the ability to shake things up in Washington was the most important criterion for their votes. The trouble for Dean was that this constituted only about a third of the voting population, according to the exit polls.
Kerry also won among voters who said that "cares about people like me," "has the right experience," and "has a positive message" were the most important candidate characteristics they took into account.
The exit poll gave voters a chance to select which of a series of issues they considered to be most important, and Kerry was the winner among those who selected most of them -- the economy, healthcare/Medicare, taxes, and education. Dean did better among those who said the war in Iraq was most important. But our weekend poll showed that the war was a distant third in priority to voters, and the exit poll showed exactly the same thing -- the war was less important to voters than the economy and healthcare/Medicare.
National Reaction to Kerry's Surge
The next question of interest is how much the Feb. 3 contests in seven states (including Arizona, South Carolina, Missouri, and Oklahoma) will be affected by what transpired in New Hampshire on Tuesday. National polling conducted immediately after the Iowa caucuses -- in which Kerry scored a major win and Dean placed a disappointing third -- has already shown that Democratic primary preferences across the nation have been transformed. Where Dean had led the Democratic pack in early January by 5-13 points (depending on which poll one looked at), polling immediately after Iowa showed Kerry leading Dean by 12 or more points nationally. The question now is whether Kerry's win in New Hampshire will strengthen his lead among Democrats nationwide, or whether Dean will regain some momentum and challenge Kerry for the lead.
Of course, national poll data at this stage are important only insofar as they relate to candidate preferences in the upcoming state primaries. As noted, the national polls conducted the week between the Iowa caucuses and the New Hampshire primary detected a strong surge of support for Kerry that was also seen at the polls in New Hampshire on Tuesday. So, that is the latest best evidence that national trends in the campaign may be valuable indicators of what to expect on a state-by-state basis. Whether that continues in places like South Carolina, where Edwards will make a strong play for votes, or states where Dean has a strong and well-financed local organization, remains to be seen.
Some Historical Perspective
Gallup election trends dating back to the 1950s show that the early contests -- the Iowa caucuses and New Hampshire primary -- have only infrequently had a major impact on national support for the candidates.
In recent elections, the New Hampshire primary has not done much to change the character of the Democratic nomination nationally. But that may owe to the fact that the primary process has changed, and surprise results in New Hampshire have served to only temporarily weaken, but not derail, the candidate with the advantage in fundraising and organization (even more so in the Republican primaries, where Bob Dole in 1996 and George W. Bush in 2000 quickly rebounded from setbacks in New Hampshire). Even though Al Gore won a narrow victory over Bill Bradley in the 2000 New Hampshire Democratic primary, Gore's lead in preference among Democrats nationally did not change -- roughly two-thirds of Democrats were supporting Gore over Bradley both before and after New Hampshire.
2000 Nomination Preferences Among National
Democrats,
Before and After Feb. 1 New Hampshire Primary
|
Candidate (pct. of vote in New Hampshire) |
Pre-New Hampshire |
Post-New Hampshire |
|
% |
% |
|
|
Gore (50%) |
67 |
65 |
|
Bradley (46%) |
21 |
24 |
Likewise, in 1992, eventual nominee Bill Clinton's level of support nationally was unchanged after the New Hampshire primary, with roughly 40% of Democrats supporting his candidacy. Paul Tsongas' win in New Hampshire did help Tsongas to vault into second place, increasing his support nationwide from 9% to 31%. However, by mid-March, after Bob Kerrey and Tom Harkin dropped out of the race after their poor showings in New Hampshire, Clinton had won the support of a majority of Democrats nationally.
1992 Nomination Preferences Among National
Democrats,
Before and After Feb. 18 New Hampshire Primary
|
Candidate (pct. of vote in New Hampshire) |
Pre-New Hampshire |
Post-New Hampshire |
|
% |
% |
|
|
Bill Clinton (25%) |
42 |
41 |
|
Paul Tsongas (33%) |
9 |
31 |
|
Jerry Brown (8%) |
16 |
7 |
|
Bob Kerrey (11%) |
10 |
6 |
|
Tom Harkin (10%) |
9 |
5 |
The New Hampshire results do seem to have had an effect on some earlier nomination campaigns. The most dramatic example was in 1984. Only 2% of Democrats nationally were supporting Gary Hart for the Democratic Party's presidential nomination prior to the New Hampshire primary (this poll was also conducted before the Iowa caucuses, which Walter Mondale won by a wide margin). Hart's surprise victory in New Hampshire propelled him to a tie with Mondale among Democrats nationally. The first poll conducted after New Hampshire showed 37% of Democrats nationally supporting Mondale and 35% supporting Hart. By that poll, John Glenn's and Jesse Jackson's national support had taken a fairly big hit, as had Mondale's.
1984 Nomination Preferences Among National
Democrats,
Before and After Feb. 28 New Hampshire Primary
|
Candidate (pct. of vote in New Hampshire) |
Pre-New Hampshire |
Post-New Hampshire |
|
Walter Mondale (28%) |
49 |
37 |
|
Gary Hart (37%) |
2 |
35 |
|
Jesse Jackson (5%) |
14 |
6 |
|
John Glenn (12%) |
13 |
8 |
|
George McGovern (5%) |
3 |
1 |
|
Alan Cranston (2%) |
2 |
-- |
|
Ernest Hollings (4%) |
2 |
-- |
|
Reubin Askew (1%) |
3 |
-- |
By mid-March, just a few weeks after the New Hampshire primary, Hart had overtaken Mondale nationally by a 39% to 30% margin. The two candidates traded the lead for the next two months, but by May, Mondale had secured his place as the candidate most preferred by Democrats nationwide.
Carter is given a lot of credit for paving the modern path to the presidential nomination by contesting and winning primaries in a variety of states. In 1976, Carter was a virtual unknown before his wins in Iowa and New Hampshire, but contrary to conventional wisdom, his victories in those states did not make him the national front-runner. Prior to the Iowa and New Hampshire wins, just 5% of national Democrats supported Carter, which increased only modestly to 14% following those contests. Even after New Hampshire, Ted Kennedy and George Wallace were still preferred by more Democrats nationwide (even though Kennedy had indicated he would not run). In those days, not all candidates competed in all the primaries. In fact, only five candidates were on the ballot in New Hampshire, even though many others contested primaries in other states (Henry Jackson, George Wallace) or did not rule out a bid for the nomination until late in the campaign (Hubert Humphrey).
1976 Nomination Preferences Among National
Democrats,
Before and After Feb. 24 New Hampshire Primary
|
Candidate (pct. of vote in New Hampshire) |
Pre-New Hampshire |
Post-New Hampshire |
|
% |
% |
|
|
Ted Kennedy |
21 |
20 |
|
George Wallace (1%) |
18 |
17 |
|
Hubert Humphrey (6%) |
15 |
13 |
|
Jimmy Carter (28%) |
5 |
14 |
|
Henry Jackson (2%) |
6 |
5 |
|
Morris Udall (23%) |
2 |
3 |
|
Birch Bayh (15%) |
1 |
3 |
|
Edmund Muskie |
7 |
3 |
|
George McGovern |
5 |
3 |
|
Fred Harris (11%) |
<1 |
1 |
|
Sargent Shriver (8%) |
2 |
2 |
By early March, Carter's support had moved into the mid-20% range, tying him with Humphrey. It was not until early May that Carter finally overtook Humphrey nationally (39% to 30%); by late June, Carter had gained support of a majority of Democrats nationally.
The 1972 campaign for the Democratic nomination was contested in a similar environment to the 1976 campaign. This was the first year that state primaries and caucuses largely determined the Democrats' presidential nominee. Prior to New Hampshire, Edmund Muskie held a slim lead over Ted Kennedy (who was not a declared candidate) and Hubert Humphrey in national support. George McGovern received only 6% national support despite a strong second-place finish to Muskie in Iowa. McGovern's second-place finish in New Hampshire also did little to boost his national fortunes. Not until May did McGovern catch Humphrey and Wallace, even though his second-place finishes in Iowa and New Hampshire were viewed as victories. In the final national preference poll in late June, the candidates were still very close (McGovern 30%, Humphrey 27%, Wallace 25%). McGovern went on to win the nomination in a vote at the convention.
1972 Nomination Preferences Among National
Democrats,
Before and After March 7 New Hampshire Primary
|
Candidate (pct. of vote in New Hampshire) |
Pre-New Hampshire |
Post-New Hampshire |
|
% |
% |
|
|
George McGovern (37%) |
6 |
7 |
|
Edmund Muskie (46%) |
20 |
19 |
|
Hubert Humphrey |
16 |
24 |
|
George Wallace |
14 |
20 |
|
Ted Kennedy |
17 |
-- |
|
John Lindsay |
5 |
6 |
|
Henry Jackson |
4 |
6 |
|
Eugene McCarthy |
4 |
6 |
|
Sam Yorty (6%) |
1 |
1 |
The effects of the 1988 New Hampshire primary are hard to read, because Gallup polled infrequently during the 1988 nomination campaign (the post-New Hampshire poll was conducted nearly a month after the primary, and followed many other states' primaries, including the Super Tuesday primaries in southern states). But by the time March rolled around, Michael Dukakis had emerged as the leader. Dukakis had a strong win in New Hampshire, but he may have emerged as the leader in part because former national front-runner Gary Hart dropped out after a disappointing finish there.
1988 Nomination Preferences Among National
Democrats,
Before and After Feb. 16 New Hampshire Primary
|
Candidate (pct. of vote in New Hampshire) |
Pre-New Hampshire |
Post-New Hampshire |
|
% |
% |
|
|
Michael Dukakis (36%) |
16 |
32 |
|
Gary Hart (4%) |
23 |
-- |
|
Jesse Jackson (8%) |
15 |
23 |
|
Paul Simon (17%) |
9 |
5 |
|
Al Gore (7%) |
6 |
17 |
|
Dick Gephardt (20%) |
9 |
10 |
|
Bruce Babbitt (5%) |
4 |
-- |
Even though there was a contest between Carter and Kennedy, Gallup did not ask the ballot question at all in 1980. Similarly, in 1968, Gallup did not ask about Democratic nomination preferences prior to New Hampshire. Eugene McCarthy's strong showing in New Hampshire is credited with leading to Lyndon Johnson's decision not to seek re-election (Johnson defeated McCarthy, 50% to 41%, even though Johnson's name did not appear on the ballot).
Gallup will update national preferences in this year's nomination campaign and the results will be released early next week.
Nearly 4 in 10 Americans Don't Know John Kerry
One effect of the Iowa and New Hampshire victories for Kerry will undoubtedly be to increase his name identification across the country. According to Gallup polling conducted in early January, prior to the Iowa caucuses and the New Hampshire primary, a large number of Americans did not have an opinion of John Kerry at all, and those who did were evenly divided in their opinions of him. Thirty-one percent of Americans said they had a favorable opinion of Kerry, while 32% had an unfavorable opinion of him. Nearly 4 in 10 Americans, 37%, said they had either never heard of Kerry or did not have an opinion about him.
At that point, in early January, Dean and Clark were about as well known as Kerry. Joe Lieberman was somewhat better known, while John Edwards was less well-known. We assume most of these candidates are better known now than they were even a few weeks ago; forthcoming Gallup Polls will help us verify that assumption.
|
Favorability of Democrats Candidates For
President
|
|||
|
|
|
Never |
|
|
% |
% |
% |
|
|
Joe Lieberman |
38 |
37 |
25 |
|
Wesley Clark |
37 |
26 |
37 |
|
John Kerry |
31 |
32 |
37 |
|
Howard Dean |
28 |
39 |
33 |
|
John Edwards |
24 |
24 |
52 |
