Share story

WASHINGTON (AP) — The congressional investigation of the 2012 attack that killed four Americans in Benghazi, Libya, is not a burning issue for the public, except for Republicans, a new Associated Press-GfK poll finds.

But overall, Americans are more likely to view the investigation as justified rather than as a political attack on former Secretary of State Hillary Rodham Clinton, who testifies before a congressional committee Thursday.

Here are some things to know about public opinion on Clinton and Benghazi from the new AP-GfK poll.



Many Americans just don’t have an opinion about Clinton’s handling of the investigation. Four in 10 say they neither approve nor disapprove of how she has answered questions about the attack, while 20 percent approve and 37 percent disapprove.

Republicans in the poll overwhelmingly dislike how Clinton has handled questions on the attack — 71 percent disapprove, including 57 percent who strongly disapprove.

But about half of both Democrats and independents don’t have strong enough feelings on the issue to say either way.

Democrats who do have an opinion are more likely to approve than disapprove, 39 percent to 12 percent. Independents are more likely to disapprove than approve, 29 percent to 11 percent.



Despite the lack of strong opinions about Clinton’s handling of the issue, Americans don’t mind the congressional investigation. They are more likely to view it as a justified attempt to try to find out what happened than as a politically motivated attack on Clinton, 52 percent to 44 percent.

On this, too, there is a major partisan divide.

Among Republicans, 83 percent consider the investigation justified and 15 percent consider it politically motivated. Among Democrats, 71 percent consider it politically motivated and just 28 percent justified.



Clinton is also likely to face questions Thursday about her use of a private email account and server, rather than a government address, while she was secretary of state.

Americans are divided on what they think is the most likely reason Clinton did that. Just under half (47 percent) think it’s because it was more convenient for her to have only one email address, about the same as the 49 percent who say it’s because she wanted to shield her emails from government transparency laws.

On this, too, Americans appear to view the issue through a partisan lens. Seventy-one percent of Democrats believe Clinton is more likely to have done it for convenience, while 73 percent of Republicans suspect she wanted to shield her emails from scrutiny. Nearly 9 in 10 Republicans supporters of the tea party movement think she was trying to shield herself from scrutiny.

Independents are slightly more likely to say she wanted to avoid government transparency laws than that she did it for convenience, 48 percent to 40 percent.

The poll finds that perceptions about Clinton’s honesty remain a weak point. Two-thirds of Americans describe Clinton as only slightly or not at all honest, while a third say she is very or somewhat honest.



Americans are slightly more likely to view Clinton’s use of a private email server as a minor problem than a major problem, 37 percent to 33 percent. An additional 26 percent say it’s not a problem at all.

Among Democrats, 10 percent call it a major problem, 52 percent a minor problem and 37 percent not a problem.

Among Republicans, 64 percent call it a major problem, 24 percent a minor problem and 11 percent not a problem.



Just 21 percent of Americans say they’re following news about Clinton using a private email server extremely or very closely. An additional 33 percent say they’re following the subject somewhat closely, and 43 percent say they’re not following too closely or not following at all.

Partisans on both sides are more likely than independents to say they’re following very closely. About a quarter of both Democrats and Republicans, but less than 10 percent of independents, say so.


The AP-GfK Poll of 1,027 adults was conducted online Oct. 15 to Oct. 19, using a sample drawn from GfK’s probability-based KnowledgePanel, which is designed to be representative of the U.S. population. The margin of sampling error for all respondents is plus or minus 3.3 percentage points.

Respondents were first selected randomly using telephone or mail survey methods, and later interviewed online. People selected for KnowledgePanel who didn’t otherwise have access to the Internet were provided access at no cost to them.



Poll results:


Follow Emily Swanson on Twitter at