November 20, 2018

Interpreting the 2018 Election



What are the implications of the 2018 election results? Julia Azari and Rachel Bitecofer are two political scientists who followed it closely and know how it compares to prior cycles. Azari is an election analyst and party scholar who finds that politicians claim electoral mandates for action based on the results of elections. We talk about early interpretations of 2018 as a referendum on Trump and why we simplify election results with stories. Bitecofer is an election analyst and a forecaster of the 2018 election who finds that demographics and partisanship are now destiny. We talk about why negative partisanship makes election results easier to foresee as partisans choose clear sides but shift turnout.

Studies: “Delivering the People’s Message” and “The Unprecedented 2016 Presidential Election
Interviews: Julia Azari, Marquette University and Rachel Bitecofer, Christopher Newport University

Transcript

Grossmann: This week on Political Research Digest, interpreting the 2018 elections. For the Niskaen Center, I’m Matt Grossmann. On this special edition, we review the implications of the 2018 midterms. I’m joined by two political scientists who both followed it closely and know how it compares to other cycles. Julia Azari of Marquette University is the author of Delivering the People’s Message, about how politicians claim electoral mandate for action based on the results of elections. We talk about early interpretations of 2018 as a referendum on Trump and why we simplify election results with stories. Rachel Bitecofer of the Wason Center at Christopher Newport University, is the author of the Unprecedented 2016 Presidential Election, and was a forecaster of the 2018 election.

We talk about why negative partisanship makes election results easier to foresee as partisan choose clear sides, but shift turnout. So what did we learn from the 2018 midterms? Azari says it’s difficult to have a big swing when partisans are so dug in to their side.

Azari: What we learned, number one, identity is on the ballot, but not always how you think. The other one is that this was an election of national sorting, but again, not in a completely simple or straightforward way. We saw things happened that consolidated the major geographical and partisan divisions within the country, and told us more about what those are. Those are the major takeaways for me, and I’ve been thinking a lot and I put this question to my students yesterday. What does a change election in a divided country? Is that even possible? Was this a change election? Was it not? I think that that’s to some degree an open question. Those are the big things that I’m looking at.

Grossmann: She says the early interpretations this year have focused on looking ahead to 2020.

Azari: I’ve been tracking this closely and I’ve been trying to think about how to compare this with previous midterms, and in terms of the volume of interpretation, and in terms of the immediate translation from last Tuesday’s results into some kind of story about what the Democrats should do in 2020. It seems to me like the 2020 campaign is looming even larger over this than 2012 did over Obama after the 2010 midterm, or other previous elections like that.

The interpretations I’m seeing, I see a lot of interpretations about the degree to which this was a referendum on Trump. I think this is really interesting, because that is not … If you were to just look at the data, you wouldn’t really need a referendum on Trump. You just look at raw data and you say, “Okay, here’s a president who is elected in a surprising election, but lost the popular vote.” He’s been remarkably unpopular, given the state of the economy. He was unpopular as a candidate. His opponent was unpopular, too, and it’s entirely possible she’d be unpopular if she were the president, but you look at that, you look at the response to some of the administration’s top-line policy, which have particularly in the area of immigration, I think has been especially interesting because they’ve lost even high profile Republicans.

You say, “Okay, let’s have a referendum on this president, and it’ll come out, and it’ll be split.” That doesn’t actually make any sense. You don’t really need a referendum on Trump. You have a lot of different pieces of data to tell you this is a president who is divisive. This is a president who’s unpopular. This is a president … When I say divisive, I mean not just that his behavior can be divisive, but that this is someone who’s polarizing. Some people feel very strongly in favor. Republicans overwhelmingly favor. Democrats, I think his approval among Democrats has been in the single digits since September. You don’t really need an election to read the tea leaves and tell you where the nation is.

Grossmann: Azari’s book, Delivering the People’s Message, focuses on how presidents claim mandates in a polarized era.

Azari: When presidents have been in a situation where they’ve treid to expand their power into new realms, they try to push at the boundaries of their power, they’ve often drawn on the concept of a presidential mandate to do so, and this goes back into the 1830s and even before. In that sense, it’s developed into a very distinctly presidential concept, because that’s something presidents have done. Particularly, I note that in the post Watergate era, presidents are always in that moment of pushing at the boundaries of their power because there’s always someone pushing back at the very legitimacy of executive power, and the concept of expansive modern executive power is always politically contested.

One way for presidents to respond to that is to say, “Well I’m doing what I was elected to do.” What’s exacerbated that is the polarization of that same era. One of the things I found in my book … My book came out in 2014. I sent my final stuff in in 2013, but didn’t include Obama’s second term, so Obama’s first term is my last case. But Obama and George W. Bush in his second term really demonstrate a turning point and how much they rely on this rhetoric just as a sheer proportion of the number of speeches that they give, and messages that they issue.

They make reference to the election, to the rejection of our opponents, to following our campaign promises. That’s one of several pivotal turning points in the ways the presidents use elections to bolster what they do, and that’s at a time when the country’s becoming hyper polarized.

Grossmann: Donald Trump has been unabashed and repeatedly claiming a mandate.

Azari: He talks quite frequently about his election victory, about how he won the electoral college, what he ran on. One of the things that I found when I started looking more closely at things like the radio addresses is that actually he used a lot of mandate rhetoric that’s pretty toned down relative to what he says on Twitter, but actually it’s a high volume of what we would see from other presidents which is like, “I am responding to the people who elected me. I am responding to my campaign promise.” He’d especially trot this out talking about immigration, so Trump really fits the pattern.

In some way this is … I don’t know how many other social scientists have run into this problem where you publish a book and then your best case happens two years after you write the book, the most interesting and most illustrative of the dynamics that you’re talking about. Some of that is compensating for legitimacy challenges. I think it is fair that Trump has faced a lot of pushback from members of the media. He’s unpopular, which is something else I found to be correlated with mandate claims. That’s real. When the defensiveness that his rhetoric displays is in response to real challenges, using mandate rhetoric in that defensive posture is very much predicted by the decades of presidential rhetoric that come before.

Grossmann: His first take on 2018 was decidedly self-centered, but Azari says it does have precedent.

Azari: The people I campaign for, the whole … the line about Mia Love and Whatney Love, that’s not standard presidential discourse. Usually these ego moments, and come out maybe in private exchanges. They don’t make their way into the first press conference. I will say it’s true. Obama talked with a shellacking in 2010. I also have a paper with Justin Vaughn. It’s floating around. It got accepted somewhere. I think it’s somewhere in the publication backlog where he looked at Obama’s rhetoric after 2014, and his rhetoric after 2014 was not totally in the shellacking vain.

He did say things like, “Look, the 2012 election still counts.” It’s not unusual for presidents to have a bad midterm and to say something along those lines. The presidential election still counts. I’m still president. We’re going to work together. I think if I’m remembering right, Clinton said things after ’94 that were very much, “The message we have is against gridlock. What this was was a rejection of gridlock. Not a rejection of me, but a rejection of the overall problematic nature of partisanship, and so what this really is is a call to bipartisanship.”

Rhetoric is rhetoric. A lot of that is trying to polish, I won’t use a crude expression about what we might be polishing after a midterm election, but trying to polish a pretty bad result. That’s on the one hand relatively transparent, and on the other hand, what are presidents going to say? They’re not wrong. They’re still president. The previous election still counts, but I will say that while Trump’s personally immodest rhetoric is new, it’s not uncommon for presidents to push back a little bit on that interpretation of a midterm that this was just a total rejection of what they stood for.

Grossmann: Usually the collective interpretation of an election takes a while to form, but coalesces by the spring following it.

Azari: The stories often are, there’s a whole bunch of stories immediately after an election, and then they narrow into one clearer narrative. What I found was that presidents typically stop talking about their election somewhere between April and June. The logic of the thing moved on. Typically at that point, they had pushed the agenda items they wanted to push early on. Sometimes it was successful, sometimes it wasn’t. Going into that first summer of being president, right outside the first 100 days, that becomes a different political environment.

The president’s relationship with Congress is starting to take shape. New events have happened, and so the political conversation has shifted and moved on. That’s honestly been less true with Trump. One of the things that I notice about him was he kept talking about his election victory well into 2017 and 2018.

Grossmann: Midterm interpretations are sometimes less clear and can evolve based on the president’s later performance.

Azari: In 2006, 2014, and 1994 are an interesting group of elections to compare for a look at recent midterms. On the one hand, ’94 is another one that other mandate scholars talk about as unexpected, and it also had a clear national … It was really in some ways the first truly nationalized congressional election where Newt Gingrich was, for lack of a better word, a congressional ringleader of the conservative movement in the House. You had the Contract with America, you had some clear policy agendas like, “Okay, what is this about? It’s about this.”

It just made for a narrative. 2006, I think, though, was not less clear in the policy issue that it was about. 2006 was very much about a turn in public opinion in the War in Iraq. You can see that in the way that Bush even responded to it. It’s at points defensive and at points making actual changes. I think that was clearly the issue. 2014 may be not as clear what the issue was, but the other difference is that you need more of an explanation when it’s an earlier midterm as a presidency goes on if that president gets reelected. You n a story what that means.

To some degree, ‘6 was like, “Okay, this is a repudiation of the Iraq War,” but at this point, Bush is only-

This is a repudiation of the Iraq War, but at this point Bush is only going to be in office for two more years, and it’s the same thing with 2014. I could really feel the end of Obama’s presidency coming about when he gave his State of the Union Address in 2015, and essentially went back to all of these promises he had made in his inaugural and even in his 2004 DNC speech about finding common ground. He kind of said, “Well, maybe we can’t find common ground, maybe we can at least agree on this other thing” and you kind of feel these presidencies winding down.

Grossmann: In 2018 we’re immediately asking if it’s a wave election with unclear definitions.

Azari: A wave election is really an interesting phrase. It dates back earlier than this, but I remember it being used a lot in 2014. And to some degree here’s what I think about a wave it’s sort of similar to a mandate, if you have to ask, it probably wasn’t. But the most common definition … Let’s talk about what’s fairly obvious in front of us. The most common definition I think of the wave is that Democrats or Republicans, or whoever did well in areas you wouldn’t expect, and did well in places where they wouldn’t normally have much business doing well.

That’s a tall order in a divided country, and it also fails to take into account the highly varied and dynamic nature of partisan trends.

Grossmann: Azari says elections often feature blunt demographic interpretations like this year’s focus on suburban educated women.

Azari: Leading up to 1996 … I recently had to explain this to my undergraduates because they had never heard the phrase soccer mom and they thought I was out of my mind. But in 1996 you couldn’t go anywhere without hearing about soccer moms, it was never clear if that was really a thing. We probably will hear about that, and I think demographic interpretations of elections are likely to be an emerging part of this story.

This is, I think, dangerous because election narratives tend to pretty blunt. And you do see some election narratives on the losing side. In the ’70s and ’80s from Democrats that essentially are about we focused too much on “special groups”, meaning women and minorities and not enough of on ordinary Americans. You kind of see echoes of in 2016, and I think that, that’s … Election narratives, as they’re cast, are very bad at dealing the complexity of the way that people’s identities intersect. So, we got women as a group, but women have different educational levels, they have different racial identities, they have different geographical faces, that’s kind of I think dangerous particularly when you have a situation in which you have a growing non white population, but still a majority white population. You have to really be careful about how you talk about constructing majorities. And so, that’s something I would encourage people as they’re kind of spinning out election narratives is to be a little bit nuanced about that kind of complexity, and be careful about how use words like ordinary, or special groups.

Grossmann: But this years elections did not have obvious policy implications.

Azari: There is research that shows that politicians do what they say they’re going to do, and that’s important. My sense is that the election narrative often takes on a totally different issue than the one that the campaign actually emphasized. That might be harder in an era of more nationalized congressional elections, and in an era in which ads go viral, and so like everyone is seeing all the ads from all around the country, and it’s so clear, at least to people who are thinking about this stuff, it’s so clear that Democratic candidates emphasized healthcare. I think that if they didn’t take that up that would probably end badly for them.

I do think from a policy perspective we’re looking at a split chamber, and a divided government under those conditions, we’re probably not looking at a terrifically policy productive period, and the stuff that will be on the agenda will be stuff that can fly under the partisan radar. So, it would surprising to me if there was a real shift on healthcare, but maybe.

Grossmann: Even though everyone is looking ahead to 2020, mid-term elections do not usually tell us much about the next nominee.

Azari: I think it’s really telling that after … So, 2010 you have this Tea Party movement that helps bring the Republicans to majority in the House, and also takes down encumbrance, it takes down like long time politicians, and never the less in 2012 the Republicans nominate Mitt Romney. I wouldn’t say a terribly good example of a mobilizing candidate, and there’s other people that they maybe could have recruited for that, but they don’t, and you see that in 2012. And you see a similar dynamic from ’94 to ’96 where maybe you want kind of new conservative firebrand, but instead it’s Bob Dole. Same Bob Dole from 20 years ago.

So, I think in some ways parties are really … I’m running with this right now, if parties are kind of intrinsically conservative institutions they’re risk adverse, their coalitions don’t move very fast.

Grossmann: Meanwhile, Rachel Bitecofer has been trying to predict the 2020 elections with her own race ratings and model based on turn out surges from negative partisanship.

Bitecofer: My model is completely different. What I’m doing in my model is I’m looking at each district, and I’m using a couple of key variables, it’s very partisan harmonious, partisan voter index, the rate of college education, the diversity and the urbanicity of the district, and using that to get an estimated white hat for Democratic party to party vote share. And then I used that to isolate districts that would be most vulnerable to a surge of Democratic voter turnout, which under my theory of negative partisanship, and that’s what my model’s called The Negative Partisanship Model, we were going to see a new electorate emerge, especially in suburban areas with high concentrations of college educated voters.

I wanted to anticipate which ones would be competitive, and in my post-op, which should be up now, it will show just when we started off in July first I had isolated 20 or 30 districts that really had a probability of producing a turnout surge for Democrats, and the other handicappers and forecasts. I’ve heard that FiveThirtyEight wasn’t out yet, but there was a new kid on the block The Crosstab done by Elliot Morris that had probabilities for districts, and they were still debating can they make it to 20, 30, and I was coming out and saying, “No. No. No. It’s probably going to be close to 40 if not more.” And the reason why I was able to account for what was going to happen in that composition, the demographic composition, or the electorate via my model.

Grossmann: She was inspired by thinking the other models were too timid after 2016.

Bitecofer: What inspired me, I think, was coming out of the November election with the upset for Clinton. I just kind of expected, I thought, like the political class to start talking about Democrats being fired up, and just kind of assume this wave was going to be coming. And then, instead, I saw that people were very tepid about it. And I think a lot of that had to do with the fact that people were too bold, maybe, about Clinton’s prospects, and the classic human response to something like that is to take it way to the other side.

Through the 2017 … I’m in Virginia. I’d be polling in Virginia for the 2017 cycle I found that when I was doing analysis, or media events, or whatever I was the only one saying, “No. No. No. This is going to hit the Democrat pretty heavily.” Everyone else is like, “No. No. It always ends up close.” I was like, “Yeah, but what we don’t appreciate, I think, is that this data journalism era that has brought like the lay person, the political lay junkie, into the process is really grounded in the last eight years, and because of that the assumptions that it operates on is assumptions that were built under the Obama presidency.” And given that we both studied political polarization, I think, probably have a special insight to just how much American politics has changed over the last couple of decades. And it just seemed obvious to me that polarization was going to have a massive effect on these mid-terms, and on last years Virginia elections. And so, when nobody else was doing that lane I said, “You know what? I’m just going to do it myself.”

Grossmann: Bitecofer uses mostly quantitative data, but does add some qualitative ratings of candidates strength.

Bitecofer: I added a qualitative element just like FiveThirtyEight does. They have their initial aggregate model on generic ballot and what have you, and then they add in data for polling. I don’t use polling, but they use polling, and fundraising, and candidate quality, and things like that to make ultimate judgements that acquire, because they were cases like Virginia seventh, I think is a great example of this.

In the case of Virginia seven that was something that no one else had listed as will flip to Democrats on July one, and I was very bullish on that district partially because of the changing northern end of the district, which is a growing suburb of rich men, which is a very Democratic strong Democratic stronghold here in Virginia, and not only that though … So, there’s that potential for that suburban surge from not just white college educated women but college educated voters in general, but what made that district enough to get it over the top … Because my two party vote share, I can’t remember for sure, but it was probably like 47%, and that’s obviously not 15. So, what made me handicap it as a will flip was the candidate. The Democratic nominee was superb. She was immensely politically talented, and it was clear that she was going to be a contender. Going through to Democrats … I mean, they were raising more than a million dollars in like more than 60 districts. It’s just an astounding data point that you don’t normally see in these House elections.

Another person that I qualitatively pushed beyond my model for was Beto O’Rourke. Texas, as a competitive state, was one of the things I got the most pushback for when I released on July one, because I said, “Oh, and by the way, Texas is going to be a toss up. Yes. Texas.” Even people that really liked what I did, like Larry Sabato, were like, “I like what you did, but Texas?” And I never said it was going to flip, but I said it was going to get within that high 40 range, and be competitive, and ultimately that’s exactly what happened.

Bitecofer: So, I don’t know exactly how I’m going to handle the qualitative aspect when I go for academic publication, because the role model was great on the seat chair, but I think it’s an important element.

Grossmann: Dee was one of the few to foresee that Democrats might not need a huge national vote share to pick up major gains in House seats.

Dee: Beause it’s based on this concept of what Democratic limitation was in the suburbs, the whole time with the analysis that were coming out about, “Oh, they need to be at 14 points to get 30 seats.” I wasn’t buying it, and the reason why is that, yes, Democratic voters are concentrated in these metro areas, and there’s that structural disadvantage, and of course the gerrymandering on top, but the basis of that assumes a certain level of competition between the parties and the suburbs. And what I was arguing fundamentally … And I think this is a key point that separates me from other people, I mean, not just in the forecasting community, but probably almost everybody in the analyst community, is that I’m not arguing Democrats had a banner day, because they went out there and swayed Republican soccer mom-

Bitecofer: There are and swayed Republican soccer moms over from the Republican party because they’re uncomfortable with Trump, though I’m sure there are some cases that fit that description. What I’m arguing that is driving the change is latent democrats, women, primarily, who didn’t vote or maybe voted only casually and were really given a wake up call, I think, in November on that election when Trump won and the last two years underneath his leadership. And they’re agitated. So I said, these latent democrats that sit out these off year elections, now they’re running around, forming grassroots, organizing groups and …

Grossmann: She says elections have changed in an era of negative partisanship and polarization.

Bitecofer: What I my theory is arguing is that not only can we explain this election, we can explain the last two, even the last four, all the way back to 2006, really, where polarization finally starts to peek up in the public. And Mo Fiorina’s famous book, Culture War, purports to say, look, Americans are totally fine. Well, the data set that he’s using for that is just when the public starts to kind of lose their mind and my dissertation looked at that, too. So you see this big change in the public and in elites coming after 2006. That’s when we start to see these like Jekyll and Hyde outcomes.

As I move forward to the academic book version of this, it’s gonna be looking backwards at this idea of negative partisanship and who is motivated and demobilized by in. And then, in terms of 2020 and 2019, here in Virginia, since I live in the best place on earth, it’s always election year. My model’s gonna be used to predict the outcome of next year’s state legislative races and the 2020 Presidential election.

And then, beyond that, assuming we get a seat change in Washington, I would expect the negative partisanship variable will be helping Republicans and hurting Democrats. So, I think this is a new way of thinking about voting behavior. It pushes back at some of the most long-standing assumptions about how people behave electorally. It really pushes heavily back on the idea that we have a persuadable electorate. That these midterm effects are just purely a product of independents moving. I mean, definitely independents move in midterm effects. But what I think what’s understudied and undervalued is the base.

Grossmann: Bitecofer did miss a couple of races because it’s hard to gauge qualitative differences in candidates.

Bitecofer: My model did really, really well. Especially at picking out districts that were not on other people’s radar, even up to election day. But, there are two districts that I got wrong. Kentucky 6 is one of them. Kentucky 6 is a district with a history of flipping back and forth.

Amy McGrath ran a great campaign she raised like $6 million. And the party and committees were giving her support, so I anticipated they must be seeing something on the ground there that gave them a good sense that she could win. And then, she ultimately didn’t even come close to winning that. And I’ve had that district as will flip from day one, July 1. So that’s obviously a major blow call for me.

And then, the other one is Virginia’s second. And, Virginia’s second, I was a victim of my own knowledge. I am connected to that district, so I know a lot of the people involved on both sides. And I let that cloud my judgment. So my model had it as a very likely Democratic pick up. And I qualitatively overwrote it to push it, lean Republican based on not only other people’s survey data, but mine, which I trust a lot more. Very few polls, but what I do is high quality and it showed Scott Taylor up by … outside of the margin of the error. So, I just … I missed that one.

Grossmann: She also saw the Senate races not going Democrats way for the same basic reasons.

Bitecofer: I said, look, Democrats are facing a hostile map. And the evidence of a blue wave and negative partisanship is gonna be primarily viewable in things that don’t happen. By that, I mean, the Midwest not being the battleground. Those states were written off and seemed very obvious to be either Democratic pick ups or safe Democratic seats pretty early on. And the battlegrounds had moved to places that it should not be in. Georgia, Texas, Arizona. Right? Factor that Arizona and Nevada went to the Democrats, I anticipated based on the level of diversity, the tendency of those … especially of Nevada to kind of break weight for Democrats.

But, yeah, the Senate map, everybody was guessing. Everybody. There was nobody who had a good forecast for that. And I think, ultimately, the state level PBI scores that I created for my forecast ended up being extremely predictive. I just followed the Cook Methodology to create those PBI scores for the states. And in doing so, found that North Dakota, for instance, had a 14 … R plus 14 PBI score. As soon as I saw that, I was like, oh, there’s no way Heitkamp is gonna be able to overcome that no matter how much incumbency helps, no matter how much negative partisanship helps, because there’s no urban suburban … there’s not a really large urban or suburban population for her to get a surge out of.

And then, Tennessee, everyone was, oh, this is gonna be competitive. And the reason they think it’s competitive is they think that Republicans could break and vote for Bredesen because it happened before or while the Southern realignment was still in progress. But that state, too, had a Republican score of 14 points. And I just knew that wasn’t gonna be overcome. And I think the exit poll data bears this out. I mean, when we look through the exit poll data, the only person … there’s three people in the country that are able to get votes from the other party in a significant way. That’s Joe Mansion of West Virginia, and Larry Hogan in Maryland, and the Governor of Massachusetts. Those are the three states that have a long history of doing that.

Grossmann: She’s bullish on Democrats for 2020, pointing to their strength in the Midwest.

Bitecofer: I think we can’t understate the importance of Scott Walker’s loss in Wisconsin. Scott Walker survived a recall, two elections. The fact that he lost in an off year, midterm cycle is a real testament to how riled up Democratic voters are, even though it was a close loss, he still lost. And then, the fact that Ohio Senate, Michigan Senate, Pennsylvania Senate, those races were not competitive. And, that didn’t surprise me at all.

And the blue wall, for me, in 2020, based on my model, is looking pretty solid for Democrats, assuming that they don’t come out of the nomination process too disadvantaged in that area. I think there’s many candidates that will be fine in the Midwest, and maybe there’s a couple that will make it a little bit harder.

But the reason why I am much more bullish on the Midwest for Democrats than other analysts is because I wrote a book on the 2016 election, and in it, I show that, actually, the Midwest … this idea that the Midwest was, oh, Trump, populism, we like him. We’re gonna cross over and jump ship. And I do know that there’s people that did that. They’re Obama Trump voters for sure. But, when you look at the actual electoral results in those states, the story of what happened there is pretty obvious. What happened was white, idealistic, mostly Millennial young people who were Sanders supporters in the Primary either sat out or voted a protest ballot. And in Wisconsin, which was less than 1% between Clinton and Trump, Wisconsin’s normal two party defection rate, so people who vote but don’t vote for one of the two party nominees, is about one and a half percent. And that’s pretty standard across the country, with some exceptions like Idaho, which tend to like third party candidates more. But one and half percent. In 2016, it was six and a half percent.

And that’s true across the entire Midwest. In my analysis, I show the main predictor of defection, of a high defection rate, is being a Sanders primary state. There’s no relationship like that for Trump states versus Cruz states or what have you. A product of Democrats becoming complacent because they were in power and they didn’t have the threat of Republican policy and Republican governance …

Grossmann: Azari says the lesson was that candidates getting national energy are not necessarily those that sell well to swing voters.

Azari: The quiet people won and the charismatic people lost. That’s one of my other take aways in terms of Democratic challengers against Republic incumbents. You had Jackie Rosen, that computer programmer, Tony Evers here in Wisconsin, a former science teacher. These are not the people that made the viral ads. The people that got a lot more national attention and kind of were more telegenic, Stacey Abrams and Andrew Gillum and Beto O’Rourke did not fare as well. I know some of those races are not resolved exactly. But, to me that would seem like a sensible narrative coming out of 2018. But it’s not clear that that will be the direction the party will go, either. I think that’ll much more reflect internal coalition dynamics rather than the 2018 narrative.

Grossmann: But she warns us not to oversell any stories. We always seem to need a cleaner narrative than we get.

Azari: What’s really important here is how much we need a narrative. Elections solve certain kinds of problems. They install people in power. This election has, as I said, it sort of shifted the power dynamic to more closely reflect the partisan landscape of the country. It’s put in a number of new faces in power with new ideas, people who bring new demographics. It’s shifted, again, some Senate seats probably closer to the demographics and preferences of their states. Those are all kind of real and concrete outcomes.

Why we need a narrative so badly is … really tells us more than any election ever … any election returns ever could.

Grossmann: There’s a lot more to learn. Political Research Digest is available bi-weekly from the Niskanen Center and on iTunes. I’m your host, Matt Grossmann. Thanks to Julia Azari and Rachel Bitecofer for joining me. Please check out their books, Delivering The People’s Message and the Unprecedented 2016 Presidential Election. Then join us next time.