The House is moving toward impeachment. What are the electoral risks for Democrats in pursuit and Republicans in defense of President Trump? We learn from research on the role of the impeachment of Bill Clinton on the 1998 and 2000 elections and compare how things look today for Trump. Gary Jacobson finds that other factors overwhelmed impeachment in congressional elections, but a few Republicans may have been hurt. Irwin Morris finds that parties can work to protect their members who vote against their constituents’ opinions on impeachment. Both are expecting near-party-line votes this time with limited electoral effects, given polarized public opinion that is unlikely to move.

Studies: “Impeachment Politics in the 1998 Congressional Elections” and “Votes, Money, and the Clinton Impeachment

InterviewsGary Jacobson, University of California, San Diego; Irwin Morris, North Carolina State University


Matt Grossmann: This week on the Science of Politics, the electoral effects of impeachment. For the Niskanen Center, I’m Matt Grossmann.

As the house moves forward with impeachment, what are the electoral risks for Democrats in pursuit or Republicans in defense of President Trump? This week, we’ll learn from research on the role of impeachment of Bill Clinton in the 1998 and 2000 elections, and compare how things look today for Trump compared to Clinton and to Richard Nixon.

I talk with Gary Jacobson of the University of California San Diego, about his article on impeachment politics in the 1998 congressional elections, and his longterm tracking of public opinion and election results. Jacobson finds that other factors overwhelmed impeachment in congressional elections. Although some members did gain from bucking their party, partisan election trends are too strong for a big impact.

I also talked to Irwin Morris of the North Carolina State University, about his book, Votes, Money, And The Clinton Impeachment. He finds that parties can work to protect their members who vote against the constituents opinions on impeachment. He’s also expecting near party line votes this time, as polarization has only increased the gains for going along with party leaders.

We’ve only had a few impeachment efforts, but we may be able to learn a little bit from the last one. Jacobson says the big lesson is expect only small effects because of lots of other factors.

Gary Jacobson: There were small effects in ’98. There was also small effects in 2000. Those are legacy of impeachment. But they were effects that were overwhelmed by the usual fundamentals that drive elections, state of the economy and president’s standing with the public.

Clinton, the Democrats actually gained seats in ’98, which was the first time since the ’30s that a party had picked up seats at the midterm. It happened later under Bush, but that was the first time it had happened in a long, long time. It’s because Clinton was 66% approving, the economy was doing very well. There wasn’t a hangover of extra seats the Democrats had because he had no coattails to speak of in 1996.

And so, the outcome looks pretty much like what the standard models, based on the economy presidential approval would have predicted. And in fact, a normal outcome in a very abnormal year. It looks like you can find a few cases where position taking on impeachment might have had some effect on the outcome, but maybe not very many.

And by 2000, of course, it was not entirely forgotten, but it was superseded by events. There weren’t a lot of Democrats angry about it and and willing to take it out on Republicans because what did they have to be angry about? Impeachment failed, Clinton remained in office, they liked him being there, they got what they wanted, which was the status quo, and so they weren’t too angry.

Republicans didn’t want to dwell on impeachment either because it hadn’t been a popular success. It hadn’t been a political success in ’98.

But Trump, my guess is it’ll be pretty much the same. If you look at his numbers, they’re just astonishingly stable. Over the course of almost three years now that he’s been in The White House, his approval ratings have been pretty much flat, highly polarized, gone up a little bit among Republican, down a little bit among Democrats. They were already very low among Democrats, down in the single digits. Now we’re in the five percent range. In the 80s, sometimes in the high 80s for Republicans. But those numbers have been very, very stable in the face of all sorts of things that have gone on, that one would have thought might’ve moved the public one way or another.

Matt Grossmann: Morris says there was some effect in the 2000 elections for the Senate, but the party was largely able to defend House Republicans.

Irwin Morris: What I found was in the House, it mattered relatively low. House members that cast votes that were unpopular with their local constituents, were able to raise a sufficient amount of campaign funds to offset that unpopularity. And in the Senate, it didn’t matter a great deal. There were a few Republicans that lost very, very close races in the Senate, that had supported impeachment, in states where impeachment was not popular. And so, there did seem to be a small with effect in the Senate but in the House, there was little or no effect.

And then when a graduate student and I from Maryland went back and looked at the impact 20 years later, [Nick Muris 00:04:52] and I looked at the career trajectories for individuals who were involved in the impeachment proceedings, either in the House and the Senate. And there was relatively little impact, so voting against your party didn’t seem to be that big a deal. Voting with your party. There just really wasn’t much of a substantial impact, as far as career development for individuals who were there at the time and participated in either the impeachment or in the in the Senate here.

Matt Grossmann: Jacobson agrees there could have been some effects in 2000. The parties were able to defend vulnerable members.

Gary Jacobson: It’s possible, but I think it’s part of a broader dynamic, in which seats that are held by a party, that are at risk for any reason, attract lots of money from the party these days, more so now than there was in 1998. But the idea is that parties would invest in their marginal seats, and try to hold onto their at risk incumbents, regardless of the source of the risk is standard. It’s what you would expect.

Whether or not that money was decisive in keeping those seats, I think is an entirely different question. Certainly, that doesn’t hurt, but it’s really hard to demonstrate that spending money like that is a decisive factor. It is true that, however, those members of Congress who support their party in districts where that’s not such a good thing, now, at least these things, tend to look very well supported by party sources in my independent groups that are allied with the parties.

They’re going to pay off for taking that risk, and that may have been true in 1998 as well. The main electoral effect that I was able to discover for the impeachment, was actually in 2000, the 2000 election, where Adam Schiff defeated James Rogan. Adam Schiff, currently leading the charge for the Democrats, defeated James Rogan, who had been leading the charge for the Republicans. That was a bit of an irony there.

There were a couple of other districts where Clinton was quite popular. Brian Bilbray, down in my neighborhood was one of them. Rogan, maybe five or six Republicans could reasonably attribute losing their seats to unhappiness with impeachment a year later. But that’s a high estimate at the height.

Matt Grossmann: Potential 1998 election effects were party wide, Morris says, rather than for individual members.

Irwin Morris: None of the votes were taken before the ’98 election. And so, in fact, there was at least one Republican who had traditionally been quite liberal, lost in the ’98 election, and then proceeds to vote yes on all four impeachment articles.

I was looking at the impact of the votes themselves. And I think those looking at the impact on the ’98 election, were thinking more broadly about the impact on the party. And so, I think there’s an argument to be made that the party, the Republican party, was weakened to some extent through this impeachment process, but not dramatically.

I mean, in ’98, I believe the Senate was a wash, and I think their Democrats only gained five seats in the house. Now, a five seat gain in an off year election is not trivial, but it’s not a dramatic shift after the election of someone who had been quite a popular president.

Even in ’98, I’m not sure there’s really a significant impact of the impeachment proceedings. And, obviously, no one is actually punishing their individual member because those votes hadn’t been taken.

Matt Grossmann: Alan Abramowitz did find some effects on individual voters in 1998, but Jacobson says that might not have been enough to matter for election outcomes.

Gary Jacobson: My first thought would be, well, opinions on impeachment are somewhat endogenous. And so, people who preferred Democrats might also have decided that impeachment was a bad thing too.

On the other hand, it’s possible that you could move a few votes at the margins. You send up with the statistically see living in effect. But if you aggregate it across districts, it might not have that effect.

I noticed that in 2018, those Republicans who were somewhat less supportive of Trump in various ways, did better. But they did better only in those districts where it didn’t matter, that is districts who were competitive, districts where Democrats actually made their gains. There was no effect for Republicans on how critical or how supportive they were.

Trump, whereas in districts where Republicans were reasonably safe, those who were critical of Trump did a little bit better, so it made him a little bit safer, but it didn’t affect the overall distribution of wins and losses.

It might’ve been something like that and in 1998, as well, that is, it moved voters in districts where they weren’t very competitive, but in the competitive districts, other things are brought before the voters that overwhelmed the impeachment aspect.

Matt Grossmann: Morris found a little member affection for Clinton, based on differences and party and constituent views.

Irwin Morris: Party was, obviously, crucially important, and then when people deviated from their party, it’s almost always explainable, in terms of their own ideology. And so, 20 plus years ago, you might have some relatively more liberal Republicans or relatively more conservative Democrats, that took a stand that was inconsistent with the majority of their party. And usually, you can understand that in terms of of their previous voting pattern. That is often related to their constituents preferences, is also true, but that’s generally true.

Matt Grossmann: But the 2000 impeachment effects were minimized by party money and by time.

Irwin Morris: In the House, you don’t have a lot of swingable elections, and where you do have swingable elections, House members were in a position to fortify their positions with additional campaign funds, which they did go out gathering, then spent.

In the Senate, if you’re in a competitive election, you’re always working to raise campaign funds. And so, it was more difficult for senators who were in a difficult situation with the party wanting one thing and their constituents largely wanting something else. They were in a somewhat more difficult situation, and a small number of them lost. I think there were five that I point to in the book. One who probably, I’m thinking that was [inaudible 00:11:49], but I can’t remember. He probably was not going to win anyway.

And then there were a set of extremely close races one could attribute to an impeachment vote, but one could also attribute to a lot of other things. So like Slade Gorton lost by 2,500 votes in Washington and he did actually cast one yes vote for conviction, but only one.

And so, in those very close races, there’s some reasonable attribution to an impeachment vote, but it was almost two years earlier, and many other things could have played a role as well.

Matt Grossmann: Jacobson says there are now few with incentives to buck the party.

Gary Jacobson: The parties who currently tend to hold … In Congress, most of the members represent districts where their presidential candidate won or did very close to winning in 2016, and so you don’t as many candidates who are, frankly, vulnerable for voting one way or another on impeachment.

Collin Peterson, if one of the two were voted against it, he’s in a the district where Trump got 66 percent of the vote. He miraculously survived in 2016 and 2018, but I think his days were numbered no matter what, once Trump comes back on the top of the ticket.

Matt Grossmann: And Morris agrees, polarization makes member affiliation unlikely.

Irwin Morris: Parties are a lot stronger than they were. The attachments to the parties are stronger. Voters are more polarized, less likely to cross over and support an opposing candidate. I think members are much more fearful of primary challenges, than they were two decades ago. And I think districts, on average, tend to be safer from a partisan standpoint, than they were two decades ago.

And so, going with the party is a safer … It was a pretty safe bet 20 years ago. I think it’s a much safer bet now.

Matt Grossmann: Primary challenges are also bigger now. Democrats in 1998 were able to split the difference on Clinton, saying it was a personal matter. Republicans have to defend Trump.

Gary Jacobson: I think it’s hugely important now. I don’t think it mattered for Clinton. I don’t think it mattered … Democrats supporting Clinton were in the position of being able to argue … and, this is one of the distinctions between these three cases, that this was something that was personal. It was about sex. He lied about it. But, you know, who wouldn’t lie about sex under those circumstances?

And if they were able to rationalize supporting him by arguing that what he did was sleazy and something they didn’t want to defend, but it didn’t have to do with governing, and therefore, he shouldn’t be impeached.

Nixon, of course, it had everything to do with governing. It was obstructing justice and investigating into his campaign’s interference and break into the Democratic National Committee offices. So, it was much more of a, it’s much harder to justify.

For Trump, it’s sort of been between. On the one hand, what the democrats are going to be impeaching him for is using his office to put pressure on a foreign country, to dig up dirt on an opposition. That’s the main thing. And, that looks like a misuse of governmental power. On the other hand, republicans by and large, we used to call them data IC, decided this wasn’t an impeachable offense. That even if he did it, it wasn’t such a bad thing to do, therefore, they’re not going to support impeachment.

So, you have an issue which, wherever there’s an ambiguity, everybody can resolve the ambiguity in favor of the partisan priors. And, I think republicans are likely to do the same thing this time around, as the democrats did back in 1980’s.

Insofar as republicans really do … republicans and Congress do think that Trump crossed the line in, and kind of bullied the Ukrainians into going after Biden through his son, they’re not saying it. And, they’re not saying it because the support among the ordinary republicans is overwhelming. And, they would generate primary challenges, were they to desert him on impeachment. So in that sense, again, public opinion is dominant here. But, it’s republican public opinion that makes it kind of political suicide in almost every place I can think of where republicans are currently hold the seat. It would be suicide because they would generate a primary challenge. That primary challenge would be well funded.

Matt Grossmann: Conservative media also keeps republican voters on Trump’s side.

Gary Jacobson: Nixon did not have vocal defenders in the media about Watergate. There wasn’t this institutionalized propaganda machine that Fox News is for Trump. I think that’s instrumental on keeping republicans on his side. It would be a fascinating test, which I don’t think we’ll get to see, if the Fox News folks soured on Trump and started to let their audience know, I wonder how much of that could be peeled off. It would be a real tension between whether or not they lose their audience, or if some of the audience would change their minds. It didn’t happen.

And, Clinton’s a little bit interesting in this way, in that he got lots of criticisms in the mainstream media. He got lots of editorial attacks. He got beaten up pretty badly, and there were calls for his resignation, et cetera, et cetera. And, they fell on totally deaf ears. John [Zallar 00:17:54] had a nice paper on that, which he kind of documents this. But, it had no effect on Clinton’s support. So, for Clinton, it helped because people ignored the media.

Trump is helped by people either ignoring the mainstream media, or being able to tune into the Fox News and the other conservative media that are relentlessly, supporters of Trump.

Matt Grossmann: And, Morris says even more money is now available to support the party line.

Irwin Morris: I think 20 years ago, you had some members that were doing some, I don’t know if I would say unusual things, but they had raised money at a significant rate that in some of their cases, they had not done previously, or had not done for a number of years. Given, again, the stronger institutional, organizational parties, the money that is available to support the party line is greater than it was then. And so, members who are going with the party line will find it easier to finance their campaigns than they did 20 years ago. And, 20 years ago, they were successful at doing it, because the key was the members that were able to raise the money to save themselves were ones who largely had gone with the party against their constituent’s preferences.

I think today, there will be fewer who are going against their constituents with their votes, and those that do will have more resources available to them to win a general election.

Matt Grossmann: The broader pattern, Jacobson finds, is that opinions are much more polarized on Trump.

Gary Jacobson: In terms of their standing with the public, you go back to Nixon, the year leading up to his impeachment, his approval rating was at 26 percent. The economy was in bad shape. We were in a recession. His support among republicans averaged about 52 percent, democrats, 13 percent approval, independent 26 percent, so there’s a partisan gap of about 39 points. But the key thing here is that his standing among republicans, his approval rating was only 52 percent.

Because Clinton, a few years later, Clinton’s overall approval rating during the year leading up to his impeachment was about 64 percent. It actually went up over the course of that year. And among democrats, it was at 88 percent. Republicans at 36 percent, not zero, with a partisan gap of about 52 points.

Right now, we’ve got Trump at 41 percent, with republicans, 89 percent, [inaudible 00:20:26] as well of him as democrats did of Clinton back in ’98. But, democrats, only six percent approval. So, you have an 83 point gap between the parties. So, you have three presidents at different levels of popular approval, in very different levels of polarization of the elector.

Matt Grossmann: Even with public shifts during the proceedings, the partisan gaps have usually remained, and are now getting bigger.

Gary Jacobson: The size of the polarization is a lot, that’s been going on for a long time. It’s just a steady increase that goes from Reagan through the end of the George W. Bush administration, and it gets wider for Clinton. It gets wider again for Bush. It gets wider again for Obama, and is wider again for Trump. So, it’s been a steady progression. The way in which the public responded to the issue of impeachment reflected these differences, and during Nixon’s impeachment, there’s about a 40 point partisan gap between democrats and republicans. But, that gap doesn’t change over time, as overall, support for Nixon’s impeachment nationally, overall, rises from maybe 37 percent to about 57 percent at the time, when he resigns. So, there’s an increase of about 20 points. That increase takes place pretty much the same among democrats, republicans, and independents, although they started very different places. So at the end, 71 percent wanted Nixon out, [inaudible 00:21:57] of democrats, 31 percent of republicans, 55 percent of independents.

So, there’s a kind of party gap that is flat at about 40 points. For Clinton, nothing much changes over the course of the entire impeachment event. It kind of flat lines, support for impeachment overall flat lines in the mid 30’s, through most of that period. There’s no real trend. Republicans, it’s 63 percent, but there is some change among republicans. They go from 56 percent majority to about a 66 percent majority supporting impeachment. Democrats, no change. Independents, virtually, no change. So, the partisan gap goes from about 45 to about 55 percentage points.

So again, you have most of the change that goes on during the Clinton impeachment takes place almost exclusively among republicans, and they become a little bit more supportive of it. But, at levels that peak out around two-thirds, and maybe 70 percent in some polls.

For Trump, support for impeachment starts out really high among democrats, and [inaudible 00:23:06] elected, it’s around 70 percent. And, it stays that high until the Ukraine issue shows up in September. And, it goes up about 10 points to 82 percent. And, among republicans, it’s basically low and flat. It goes up a couple of points after Ukraine, but it’s only around 10 or 12 percent. And then, independents go up about, like, the national, which is flat at the low 40’s through late September. And now, it’s up to 47, 48 percent.

So, you have some change for Trump, but it changes almost entirely different by democrats into a second [inaudible 00:23:47]. So, you have a much more polarized situation. He is weaker in terms of public opposition to his impeachment than either Nixon was until the very end, or that Clinton ever was. On the other hand, as long as he holds onto his republican base, there will only be at best, a small popular majority in favor of impeaching him.

Matt Grossmann: Morris says Clinton did not see the drop that Nixon did, and Trump looks likely to be more like Clinton.

Irwin Morris: The biggest difference to me between the ’98 case and the previous case is the trajectory of public opinion. If you look at where Nixon was early in ’73, let’s say, before Watergate breaks, and you look at where his public standing is when he resigns, it’s a precipitous drop. Probably, I looked at the figures today. I want to say something along the lines of 40 to 45 percentages points. That obviously did not happen in Clinton’s case. And in fact, he did not see precipitous drop, and in some cases, was actually more popular, was quite popular at the end of ’98.

So, that’s a dramatic difference right off the bat. In comparison to today, I can’t imagine that sort of public opinion drop for an American president today. I mean, first of all, you’ve got to get to 65 or 70 percent, to have that drop in the first place, which is sort of hard to imagine. But, two have a sufficiently bi-partisan level of opposition to presidential behavior or presidential actions, to get that drop in public opinion is quite frankly, hard … Like I said, it’s hard for me to imagine in our day and time.

Matt Grossmann: Jacobson says the president’s party seems to defend the president, historically and now.

Gary Jacobson: The president’s party never changes much. They usually oppose impeachment. For Nixon, at the very end, the very last poll, republicans support for impeachment goes from 12 percent, up to 30 percent. But, that’s the only change you ever see in the president’s partisan. They stay on his side. All the action is on the other side, it’s in the other party’s partisans, become more supportive of impeachment as time goes by.

Again, Nixon’s a little bit of an exception. The republican co op from very low, up to the teens, and then jump to 30 at the very end. But, there is no dramatic change. And, there wasn’t a dramatic change among the democrats, either. They kind of jump up at the end, too. But, they weren’t overwhelmingly in favor of impeachment until the very end.

Matt Grossmann: But, he sees some potential effects on the presidential results this time.

Gary Jacobson: [inaudible 00:26:38] party line voting across the board on this issue, which suggests that he will be impeached, and that he will be acquitted in the Senate. The effect of this will all take place in the presidential election in 2020. And, I think that’s where the impact will be of whatever the democrats are able to bring out in public that hasn’t already been brought out in the public through the impeachment process.

I expect the election to look a lot like a combination of 2016 and 2018, where you had very close races in 2018, or 2016 in some states like Michigan and Wisconsin, and so forth where, he eked out his electoral college victory. And then, you have a kind of counter democratic, counter attack in 2018, in which they picked up seats in some of those states. And, actually had a vote statewide house, [inaudible 00:27:37] larger than what Clinton, in 2016. So, it’s going to be some combination of that.

What I think will happen is that, impeachment will contribute to a really important dynamic generated by Trump, and that is [inaudible 00:27:53] and mobilization. We had turnout in 2018 that was 10 or 12 points higher than it’s been the last several 10 or 12 points higher than it’s been over the past several decades, the highest it’s been in 100 years. I think it’ll be amplified. It’s inevitable because Trump excites strong passions on both sides, and impeachment will just reinforce those passions on both sides. I don’t think many minds are going to be changed, maybe some small proportion of the electorate that hasn’t made its mind up yet and among people who like to call themselves independents. They may still be open to moving one way or another. But I don’t think there’ll be a lot of movement on that, and so the outcome will depend largely on how effectively both sides get their passionate voters out, and I think they’re both going to be quite successful and we’re going to see the largest presidential turnout we’ve seen in many, many decades.

Matt Grossmann: And some potential for long-term effects on opinions of the Republican party.

Gary Jacobson: It’s effect will be, I think, generational. Trump is very unpopular among younger voters. I think that they’re more supportive of impeachment than older voters. This is going to have an effect in the sense that it will be a stain on the Republican party if the sentiments of younger voters, as they age, don’t change very much on this. And I don’t know… It’s possible we’ll have Trump nostalgia somewhere down the road, among that generation, but somehow I don’t see it happening yet.

Matt Grossmann: Morris fears that the details of the scandal may not matter anymore.

Irwin Morris: It’s really not clear to me that it matters. I mean, from my own standpoint, what Nixon did was institutionally devastating and had extraordinary costs associated with it, and in a way that I don’t see Clinton’s behavior in the same framework. But I mean, the key is in what is a political activity. I mean, there are no rules for evidence. There’s no appeal process if the president is removed from office. There’s nothing in the constitution that says the president must have committed a particular crime or a crime, specifically. And so even the founders realized this is a political process. That doesn’t mean it’s not a crucially important process, but it is a political process. And in 1974, you had people on both sides of the aisle who saw what the president had done as being of such a serious weight that he would need to be removed from office.

Matt Grossmann: But the Ukraine scandal is simpler for Democrats to understand.

Irwin Morris: I think it looks better for impeachment because the Ukraine scandal was not uncomplicated, but I believe more straightforward. And I think the Ukraine scandal has had more resonance with Democrats than the Russian scandal. I mean that’s, I think, the key. People who have largely supported the president have not been moved, or, if they have been moved, they’ve shifted to more support. And people who were largely opposed to the president have been moved, and they have been moved to greater opposition. At least that’s the most recent poll results that I’ve seen. And so I think that makes a difference with the Democrats because those are their constituents, and I think that was an effect and probably a more substantial effect than with the Russian investigation.

Matt Grossmann: Neither sees big electoral effects coming, but Jacobson says, “There’s more potential downside for Republicans.”

Gary Jacobson: I have yet to see any groundswell of sympathy for Trump, which is what the Democrats would have to worry. I think there’s more downside for Trump than for the Democrats on this because there will be an airing of his various transgressions in full public view. But I also think that the danger for Trump is kind of Trump fatigue, that people may get sick of all the drama. I think they already are. But sick of all the drama and, for that reason, decide that they either want to replace Trump or they’re not going to go to the polls to vote for him. They might take the edge off his support a bit.

Can imagine this kind of going on and on and people saying, “Oh, regardless of whether he should be impeached or not, we’re sick of it. Let’s change. Let’s get a change.” I think that’s the danger to Trump. It may not evolve in that direction, but I think he has a little bit more of a downside than the Democrats do for impeachment. But my basic bottom line is that it’s going to simply reinforce the lines of division that we’ve already been [inaudible 00:05:03].

Matt Grossmann: Morris says the difference this time is the variety of Republican defenses.

Irwin Morris: In some ways, this is a more complicated scenario or situation. I think in both the Clinton case and the Nixon case, even supporters were very upset and very critical of the president’s actions. I mean, in the Nixon case, very critical, and people were saying, “It’s very, very serious and is sufficiently serious that you should be removed from office.” In Clinton’s case, I think it was more… People were very critical, but among Democrats, largely not sufficiently serious. It does not rise to the standard that the constitution requires for removal. I think in the current case, you have all manner of opinions. Yes, it was serious, but it wasn’t serious enough. Yes, it was serious, and it was serious enough. No, it wasn’t serious.

I think you have some who would even say, “No, it didn’t really happen,” among the president’s supporters, that it is completely fabricated. Now, maybe I’m going a little bit too far on that, but there is a level of conflict and controversy over the details, over the characterizations of the actions, that I don’t think was true in ’74 and ’98, and so history’s evaluation of this is going to be somewhat more complicated because there isn’t really any consensus at all.

Matt Grossmann: Jacobson will be looking for any changes in attitudes across parties.

Gary Jacobson: You want to look at whether or not anybody’s mind gets changed. You want to look at whether there’s any shift in aggregate partisanship because most of the data that I look at, I want to see the breakdown by party. And that’s important, but there’s also possibilities that the distribution of partisan can shift in the course of a year or two in a way that has some kind of important consequences. So I’ll be monitoring aggregate party ID. I’ll be monitoring attitudes toward Trump, especially among Republicans. I think everything I look at, you have 90 to 95 percent of Democrats really saying negative things about Trump, and I don’t think there’s any reason to expect that to change.

Republicans are broadly supportive of him, but on some particular things, their support will drop down from 85% and 90% down to 68 percent or 75 percent. And I’ll look forward to see whether there’s any change or erosion in that. Again, I don’t expect it, barring some new revelation that even Fox News can’t explain away. But barring that, I think that there’s a possibility of some erosion of Republican support, but it’s not going to be big. It’s not going to be huge.

Matt Grossmann: Morris wants to know if there’s any divergence between constituent opinion and party leaders or if there’s no longer a trade off for following your party.

Irwin Morris: I mean, I would be looking at the same sorts of things. I think what would be interesting here is there were plenty of districts in 1998 where it was clear that the majority of the constituents did not favor the position that their incumbent’s party was taking on the issue. Now, in most cases, it was not a problem in the end for incumbents, but there were a number of districts and states where that was true. I’d be interested, today, to know how many states and congressional districts that’s actually true of. How many districts or states have a clear majority of voters or potential voters who have taken a position on the impeachment that’s inconsistent with the primary party position for their elected official? My guess is it’s a much smaller number.

Matt Grossmann: Jacobson’s sees some signs of a Republican retirement wave, mostly due to being in the minority, but perhaps from tiring of defending Trump.

Gary Jacobson: Republican retirements are beginning to look like a signal. I think it’s largely due to the fact that they’re now the minority party, and being a minority in the house is no fun whatsoever. We had a lot of Republican retirements in 2018, like 40. And we’ve got up to maybe 19 right now, and that’s a sign that they don’t want to fight anymore. This is not a bad time to get out. They’re not going to be in the majority, they don’t think, and a lot of them are… Not all of them, but some of them were going to be up against pretty serious opposition. So that suggests that they don’t think that Trump’s going to triumph and bring a Republican Congress in his wake in 2020. I haven’t seen…. Well, the nomination quality of the candidates… No idea right now because we haven’t seen the nominations take place.

I’ll be collecting that data, but I always do anyway. [inaudible 00:38:29] is the kind of year it is. I would expect high concentration of quality candidates in swing districts and District Square. The presidential vote margin is within two points of the national average, one way or the other, maybe three points from the national average. I think there’ll be a huge concentration of resources there. There was a struggle for control of the house, and I think it’ll be interesting to see how Trump plays out in those districts because they’re districts where he is more or less successful and is reasonably successful. If he just loses a couple of points in this, it’s all over for him, basically. The other thing I’ll be predicting and they’re looking for is anybody who’s running for the house should consider how they think Trump’s going to do in that district because the likelihood of inconsistent outcomes between Trump and the house vote, I think, is really small. [inaudible 00:39:31] be more than 5 or 6 percent of the districts that split in any way.

Matt Grossmann: There’s a lot more to learn. The science of politics is available biweekly from the Niskanen Center. I am your host, Matt Grossman. Thanks to Gary Jacobson and Erwin Morris for joining me. Please check out Votes, Money, and the Clinton Impeachment and Impeachment Politics in the 1998 Congressional Elections and then listen in next time.