Presidential Race Remains Close; No Convention Bounce

GALLUP NEWS SERVICE

PRINCETON, NJ -- A new CNN/USA Today/Gallup survey, conducted on Friday and Saturday following the Democratic convention in Boston, finds that the presidential race remains close, with President George W. Bush receiving 50% support among likely voters, and Massachusetts Sen. John Kerry 47%. Among the larger group of registered voters, Kerry leads 50% to 47%.

Compared with the last CNN/USA Today/Gallup survey conducted before the convention (July 19-21), this post-convention poll shows that among likely voters, Kerry's support is actually two points lower than it was pre-convention, while Bush's support is three points higher. Among registered voters, Kerry's support in the current poll is one point higher than it was pre-convention, while Bush's support is also higher, by two points (the percentage of undecideds dropped in both groups).

For neither group of voters does the slight change in voter preferences, compared with a poll conducted shortly before the convention, favor Kerry. All of these changes are within the margin of error, meaning that we cannot be 95% certain that Bush gained on Kerry after the convention. However, clearly there is no convention bounce for Kerry.

An analysis of convention bounces since 1950 shows that on average, among registered voters, the Democratic candidate received approximately a seven-point increase in support following the convention. The current poll shows only a one-point increase among registered voters. In addition, following previous Democratic conventions, the Republican candidate also lost an average of about five points, making the net effect essentially a 12-point swing in the lead. The current poll shows only a 1-point swing in the lead among registered voters, but for the Republican, not the Democrat.

In the current poll, if independent candidate Ralph Nader is included in the vote tally, Bush leads Kerry by four points among likely voters (50% to 46%), with Nader receiving 2%. Among registered voters, it's a dead heat at 47% each for Bush and Kerry, with Nader at 3%. Among both groups the percentage of voters favoring Nader and the number of undecideds is lower now than before the convention, suggesting one impact of the convention -- at least in the short term -- was to focus voters more directly on the two major party candidates.

In the three-way race, Kerry lost a point among likely voters after the convention, and he neither gained nor lost among registered voters. Bush gained four points both among likely voters and among registered voters. To the extent that there is any bounce in this three-candidate ballot, it is for Bush rather than Kerry.

Why No Democratic Bounce?

In an article on Gallup's Web site before the Democratic convention, Gallup Poll Editor in Chief Frank Newport cautioned that it was possible the average convention bounce calculated for previous conventions might not apply this year. Voters appear unusually intense this election compared with previous elections, with the number of persuadable voters only about half the total in 2000. In addition, it appears that more voters are paying closer attention to the campaign this year than in previous years, so that relatively few voters would find the convention presenting information they had not already heard.

Another difference this year from previous years is that the three major television networks cut back dramatically on the coverage of the convention, devoting a total of only three hours in prime time over a four-day period. The one hour of coverage each night on Monday, Wednesday, and Thursday came late in the evening -- 10:00 pm Eastern time, 9:00 pm Central.

Apart from these factors, it is quite possible that the aggressive response of the president and vice president immediately after the Democratic convention helped to tamp down even a slight bounce that might have occurred despite strong voter commitment and little convention coverage. Among registered voters, the Friday figures showed a lead for Kerry, while the Saturday figures -- conducted after Bush began a campaign swing in four battleground states to counter Democratic criticisms of his presidency -- showed a lead for Bush.

The CNN/USA Today/Gallup surveys over the past several months have shown surprising stability in the presidential contest, with both candidates' support levels varying within a narrow range. Kerry led by six points in early March, but then Bush led by six in mid-April. In early June, Kerry led by six, but that moved to a virtual tie, until Kerry led by five in early July. Now Bush leads by four. 

The poll also shows that only about one in eight likely voters is not firmly committed to a candidate, leaving little opportunity for a convention -- or any other events, for that matter -- to have a dramatic effect on the presidential contest. It may well be that the Republican convention scheduled for late August and early September will not produce much change in the candidates' standings either.

Survey Methods

Results are based on telephone interviews with 1,011 national adults, aged 18 and older, conducted July 30-31, 2004. For results based on the total sample of national adults, one can say with 95% confidence that the margin of sampling error is ±3 percentage points.

Results based on likely voters are based on the subsample of 763 survey respondents deemed most likely to vote in the November 2004 general election, according to a series of questions measuring current voting intentions and past voting behavior. For results based on the total sample of likely voters, one can say with 95% confidence that the margin of sampling error is ±4 percentage points. The likely voter model assumes a turnout of 55% of national adults, consistent with recent presidential elections. The likely voter sample is weighted down to match this assumption.

For results based on the sample of 916 registered voters, the maximum margin of sampling error is ±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.