What Republican voters think of candidates before the second debate

Seven Republican presidential candidates will debate Wednesday night at the Ronald Reagan Presidential Library in California. So 538, The Washington Post and Ipsos are teaming up once again to survey Republicans on their views of the candidates both before and after the debate. Former president Donald Trump still leads the pack in our poll, but many Republicans are keeping their options open.

After the first debate, our poll found more Republicans had positive reactions to Florida Gov. Ron DeSantis, former pharmaceutical executive Vivek Ramaswamy, and former South Carolina governor and United Nations ambassador Nikki Haley. Debate watchers said DeSantis and Ramaswamy performed best, while Haley saw the largest uptick in voters considering her.

The second debate could be an opportunity for one of them, or another candidate, to emerge as an alternative to Trump, especially as the former president has once again refused to join the debate. But the bar to surpass him will be high, as a Post average of September national polls found 58 percent of Republican voters support Trump, which is more than four times the support for DeSantis at 14 percent, while all other candidates are in single digits.

Below is a rundown of other key findings from our poll, including what potential Republican voters think about the candidates and key issues heading into the second debate. (Neither former Arkansas governor Asa Hutchinson nor former congressman Will Hurd qualified for the debate, but our poll still asked voters what they thought of them.) Stay tuned for data after where we will see whether Republicans who watched the debate feel differently about the candidates once they hear more from some of them.

The poll also found Trump has the most positive image among potential Republican voters, with 67 percent reporting a favorable impression of him and 30 percent an unfavorable view of him. Four other Republicans also register more positively than negatively among Republican voters: DeSantis, Haley, Ramaswamy and South Carolina Sen. Tim Scott. Haley saw the largest uptick in favorability from the last debate, of six points, while Ramaswamy saw both his favorable and unfavorable ratings increase.

There are candidates in worse positions. For instance, the image of former vice president Mike Pence remains underwater among fellow Republicans, with 39 percent holding a favorable view of him and 50 percent an unfavorable view. Former New Jersey governor Chris Christie is in even worse shape, with 20 percent holding a favorable view versus 52 percent holding an unfavorable view. Three-fourths of potential Republican voters still have no opinion about North Dakota Gov. Doug Burgum.

The top issue for potential Republican voters is inflation, with 63 percent saying it is one of the three top issues facing the country. That is followed by immigration at 48 percent and the government budget and debt at 33 percent. Meanwhile, 18 percent of Republican voters mention crime or gun violence, election fraud, extremism or polarization, and taxes. The top issues are little changed from before the previous debate.

Despite immigration ranking as the second top issue for Republican voters, there is not necessarily widespread support for some of the stances of the candidates. A question in our poll found mixed support among Republican voters for the idea from Ramaswamy and DeSantis of sending the military to fight drug cartels in Mexico. Republican voters are split on this issue with 44 percent in support of it and 46 percent opposed to it.

Notably, while the first debate highlighted the concern among younger voters about climate change, just 10 percent of Republican voters under 35 named it as one of their three top issues, which is not significantly different from the 8 percent of Republican voters who name it as one of their three top issues.

In the first debate, the candidates disagreed on whether the federal government should restrict abortions. Pence and Scott, for instance, said they favored a nationwide 15-week abortion ban, while Burgum said he was opposed to a federal restriction on abortion, even though he signed a six-week ban into law in North Dakota. DeSantis and Haley did not answer directly on the issue, but Haley repeatedly pointed to the difficulty of passing such an abortion ban in the Senate. In our poll, however, there is fairly large support among potential Republican voters for some type of nationwide abortion ban.

Fifty-seven percent of Republican voters surveyed said they support a federal ban after six weeks of pregnancy with exceptions for rape, incest or when the mother’s health is in danger. Support rises to 67 percent for a 15-week ban with the same exceptions. But more than half of Republicans, or 54 percent, said the decision of whether a woman can have an abortion should be left to the woman and her doctor, while 44 percent said it should be regulated by law. More female Republican voters said the decision should be left to the woman and her doctor at 59 percent than male Republicans at 50 percent.

Most Republican voters also said birth control pills and emergency contraception should be legal in all or most cases, while surgical abortion should be illegal. Meanwhile, slightly more Republican voters said medical abortions or abortion pills should be illegal than legal, but there is not a clear majority opinion on that issue.

Finally, nearly 7 in 10 Republican voters support impeaching President Biden, including nearly half who said they “strongly support” him being impeached. House Republicans are planning to hold their first impeachment inquiry into Biden this week, even though no clear evidence of an impeachable offense by the president has yet emerged.

The 538/Washington Post/Ipsos poll was conducted Sept. 19 to 26 among a random national sample of 5,002 U.S. adults who say they are likely to vote in the Republican primary or caucus in their state. The margin of sampling error is plus or minus 1.6 points. The sample was drawn through the Ipsos Knowledge Panel, a survey panel recruited through random address-based sampling of U.S. households.

The study was conducted in both English and Spanish, and the overall adult sample was weighted to match population benchmarks gender by age, race or ethnicity, education, census region, metropolitan status, household income and party identification.