Larry Bartels

with maybe the exception of 2008. If you look at the partisan loyalty of people within each of the two parties, it's hard to get exact figures because different surveys have different kinds of vulnerabilities to air and they vary a bit in their numbers. But typically, each party's nominee gets about 90 to 95% of the support of partisans from their party.

with maybe the exception of 2008. If you look at the partisan loyalty of people within each of the two parties, it's hard to get exact figures because different surveys have different kinds of vulnerabilities to air and they vary a bit in their numbers. But typically, each party's nominee gets about 90 to 95% of the support of partisans from their party.

The exceptions to that were cases where a nominee got probably in the high 80s, 85 to 90%. And those would be John McCain on the Republican side in 2008, who was running as the economy was melting down under an incumbent Republican president. And Hillary Clinton and Kamala Harris on the Democratic side, both running against Donald Trump.

The exceptions to that were cases where a nominee got probably in the high 80s, 85 to 90%. And those would be John McCain on the Republican side in 2008, who was running as the economy was melting down under an incumbent Republican president. And Hillary Clinton and Kamala Harris on the Democratic side, both running against Donald Trump.

Trump did a little better among men and a little worse among women than Republicans typically have. There's been a longstanding gender gap in partisanship and voting behavior. It's, I think, become somewhat strengthened recently, probably mostly as a result of people's responses to Trump's rhetoric.

Trump did a little better among men and a little worse among women than Republicans typically have. There's been a longstanding gender gap in partisanship and voting behavior. It's, I think, become somewhat strengthened recently, probably mostly as a result of people's responses to Trump's rhetoric.

which has been, I think, more outspokenly pro-male in some sense than previous candidates have been. Again, I think that's a pretty small shift at the margin. The other significant difference probably is with respect to the behavior of people with and without college degrees.

which has been, I think, more outspokenly pro-male in some sense than previous candidates have been. Again, I think that's a pretty small shift at the margin. The other significant difference probably is with respect to the behavior of people with and without college degrees.

And again, this is a kind of intensified difference recently, but one that has been growing over a significant period of time. It used to be the case that people with college educations were more Republican and people with less formal education were more Democratic. And that difference gradually closed and now has even reversed.

And again, this is a kind of intensified difference recently, but one that has been growing over a significant period of time. It used to be the case that people with college educations were more Republican and people with less formal education were more Democratic. And that difference gradually closed and now has even reversed.

Again, I think Trump's rhetoric and the nature of his appeal has something to do with that. But I think it's mostly a kind of long term response to people's understandings of the parties and what they stand for.

Again, I think Trump's rhetoric and the nature of his appeal has something to do with that. But I think it's mostly a kind of long term response to people's understandings of the parties and what they stand for.

Yes, individual candidates do matter some, and their rhetoric, I think, has some impact on the movement of these specific subgroups within the party's coalitions. But overall, the stability of partisanship and the high levels of support of partisans within each camp for their own parties and nominees seems to be pretty set, regardless of who the candidates are.

Yes, individual candidates do matter some, and their rhetoric, I think, has some impact on the movement of these specific subgroups within the party's coalitions. But overall, the stability of partisanship and the high levels of support of partisans within each camp for their own parties and nominees seems to be pretty set, regardless of who the candidates are.

Again, I think Trump is about the most dramatic, test case you could have of that proposition. A lot of people were surprised by the result of the 2016 election because he was such an unusual Republican candidate and indeed had tepid support or even opposition from many of the most prominent leaders in the Republican Party.

Again, I think Trump is about the most dramatic, test case you could have of that proposition. A lot of people were surprised by the result of the 2016 election because he was such an unusual Republican candidate and indeed had tepid support or even opposition from many of the most prominent leaders in the Republican Party.

But in spite of that, he got the overwhelming support of the Republican rank and file.

But in spite of that, he got the overwhelming support of the Republican rank and file.

Well, I think the most important systematic factor is the state of the economy. If we look historically, we see that the incumbent party does substantially better when the economy is in good shape and substantially worse when it's in bad shape.

Well, I think the most important systematic factor is the state of the economy. If we look historically, we see that the incumbent party does substantially better when the economy is in good shape and substantially worse when it's in bad shape.