Clinton maintains this edge in the general election race despite a growing perception that by using a personal email account and server while serving as secretary of state she did something wrong. About 56% say so in the new poll, up from 51% in March. About 4-in-10 (39%) now say she did not do anything wrong by using personal email. Among Democrats, the share saying she did not do anything wrong has dipped from 71% in March to 63% now, and just 37% of independents say she did not do wrong by using the personal email system.
And positive impressions of Clinton continue to fade. Among all adults, the new poll finds 44% hold a favorable view of her, 53% an unfavorable one, her most negative favorability rating since March 2001. A majority of women voters have a positive take on Clinton, 52% view her favorably, and her support among women appears to be the foundation for her general election advantages.
But the fading numbers haven't hurt her against some GOP contenders. Clinton has her biggest lead over Fiorina, topping her 53% to 43% among registered voters. She leads Bush by a nearly identical margin, 52% to 43%. And Clinton tops both Trump (51% to 45%) and Walker (52% to 46%) by 6 points each.
All of those leads are boosted by sizable gender gaps.
Against Bush, Clinton leads 59% to 37% among women, while Bush holds an advantage among men, 51% Bush to 44% Clinton. Against Fiorina, the only woman among the major candidates for the Republican Party's presidential nominations, women break 60% for Clinton to 39% for Fiorina, while men are about evenly divided, 48% for Fiorina, 46% for Clinton. The largest gender gap -- 34 points -- comes in a match-up between Clinton and Trump. Women favor Clinton by 23 points, 60% to 37%, while men break in Trump's favor by 11 points, 53% to 42%.
At the same time, Clinton's lead in the race for the Democratic nomination for president is narrowing, and the new poll suggests the best way for the former secretary of state to shift the momentum would be for Vice President Joe Biden to decide to sit this one out. Most Democrats, though, say they'd like to see Biden make a run for the White House.
Overall, 47% of Democratic and Democratic-leaning voters say they support Clinton for the party's nomination. That's down 9 points since July, and marks the first time her support has dipped below 50% in national CNN/ORC polling on the race.
At the same time, Vermont Sen. Bernie Sanders
has climbed 10 points since July and holds second place in the race with 29%. Biden follows at 14%, former Maryland Gov. Martin O'Malley
is at 2%, former Virginia Sen. Jim Webb has 1% and less than 1% back former Rhode Island Gov. Lincoln Chafee
And though support for Clinton's nomination bid has slipped, she is the candidate most trusted by Democrats on four top issues, and remains the candidate to beat in general election match-ups.
Should Biden decide not to make a run for the presidency, his supporters would largely flock to Clinton rather than Sanders, boosting her numbers. With Biden's backers re-allocated to their second choice, Clinton holds 56%, Sanders 33%, with support for O'Malley, Webb and Chafee unchanged.
Among those Democratic voters who are "extremely enthusiastic" about voting for president next year, it's already essentially a two-person race. Clinton is the choice of 50% of such voters, with Sanders jumping to 38% among that group, Biden at 6%, and O'Malley at 2%.
Still, most Democrats say they do want Biden to make a go of it: 53% of Democrats and Democratic-leaning voters say they think Biden should run, 45% that he should stay out. Even among those who currently support Clinton, 50% say they think Biden should run.
Democratic voters aren't necessarily convinced a Biden presidency would be better than a Clinton one, though: 35% say Biden would do a worse job as president than Clinton, 27% a better one and 38% say there wouldn't be any difference between the two. Those who think Biden should run are more apt to say he'd do a better job than a worse one (41% better vs. 18% worse) but a sizable 41% say there ultimately wouldn't be any difference between the two.
Fewer see Sanders as equivalent to Clinton, more say he'd do a worse job. Overall, 37% of Democratic voters think Sanders would do a worse job as president than Clinton, 31% that he'd do a better job, and 29% that there would be no difference between the two. That shifts among the party's liberal voters, 41% of whom think Sanders would do a better job than Clinton, 34% think he would do worse than Clinton and 22% that there would be no difference.
Sanders has boosted his favorability rating in the last month, 35% of adults and 58% of Democratic voters have a positive impression of the senator, that's up from 23% among adults and 36% among Democratic voters since July.
When Democratic voters are asked which candidate they trust to handle the economy, race relations, foreign policy and the income gap between rich and poor Americans, Clinton tops the list each time. Her biggest advantage comes on foreign policy, where 61% of registered Democrats say they trust Clinton over the rest of the field. Biden follows at 22%, and just 9% say Sanders is their top choice here. On the economy, 45% say Clinton would best handle it, 26% choose Sanders and 21% Biden.
Sanders fares best on the income gap, with 34% saying he would do the best job handling that issue, 7 points behind Clinton's 41%. Though the margin is wider among those with incomes under $50,000 annually, 49% of Democratic voters in that group favor Clinton to handle the income gap, 28% Sanders.
On race relations, an issue that both top Democrats have addressed directly in interactions with Black Lives Matter activists in recent weeks, Clinton tops Sanders by nearly a 2-to-1 margin, 50% choose her while 26% say they trust Sanders most, 14% Biden. Democratic voters are broadly positive toward the Black Lives Matter movement, 59% have a positive take on it, 17% unfavorable, and 24% haven't heard enough or don't have an opinion.
The CNN/ORC Poll was conducted by telephone August 13-16 among a random national sample of 1,001 adults. The sample included 897 registered voters, 358 of whom are registered voters who are Democrats or independents who lean toward the Democratic Party. For results among those Democratic voters, the margin of sampling error is plus or minus 5 percentage points. For results among registered voters it is 3.5 points, and among all adults, 3 points.