A New Theory of Political Scandal‪s‬

Political scandal is a historically defining aspect of American politics. But, there’s been very little scholarship on the political incentives that surround the production and consequences of scandals.

In a recent paper, “Political Scandal: A Theory”, our very own Will Howell and Wioletta Dziuda create a new model of political scandal that makes these incentives clear. On this episode, we discuss how these incentives should reshape the way we think about political scandals.

Listen on Apple Podcasts or wherever you enjoy podcasts.

Transcript

William Howell:

I'm Will Howell.

Wioletta Dziuda:

I'm Wioletta Dziuda.

Anthony Fowler:

I am Anthony Fowler, and this is not another politics podcast.

Anthony Fowler:

We just had the historic second impeachment of Donald Trump. We're almost at the end of the Trump presidency, which has probably been the most scandal plagued precedency in recent memory.

William Howell:

Yeah. I mean, you've got to look back to Harding to find one that has as many time scandals, but Harding wasn't impeached twice.

Anthony Fowler:

Right.

William Howell:

Right.

Anthony Fowler:

It's been difficult as a consumer of the news during the Trump presidency. There have been so many scandals that it's hard to even pay attention to them.

Tape:

One of the hallmarks of this unprecedentedly scandal written presidency is unfortunately that very serious things, very shocking things even tend to just pass by.

Anthony Fowler:

There's this there's this paper that just came out in the HAPS called political scandal. I actually just am about to have the opportunity to speak to the author.

William Howell:

Yeah, here we are.

Anthony Fowler:

The two of you have just published a paper about political scandal.

Wioletta Dziuda:

We did.

Anthony Fowler:

So we're going to divert from our typical format today. The two of you will be the guests and I will be the host and so I'll get to have my bottom line by myself, which will be fun.

William Howell:

We can't rest the microphone away from you at that moment. It will be one moment of truth at the end, and it will be yours.

Anthony Fowler:

You'll have plenty of opportunity to say your piece. But I hope this will be fun. So your paper is, we should say at the outset, it's a theoretical paper about political scandal. Why do we need a new theory of political scandal?

William Howell:

Wioletta and I came to this topic recognizing that political scandals are powerfully disruptive forces in politics, and we wanted to make sense of them. There's a naive way of thinking about scandals and it's frankly, I think the predominant way of thinking about them is that they're something like asteroids who come crashing down on earth and that the entirety of the explanation that needs to be offered is how people respond to them and that didn't seem right to us. Precisely because there's powerful incentives for people to try to conceal them or to propagate them and we wanted to make sense of those incentives. Then there also is when you see the politics of scandals play out, the discussion of misbehavior often and quickly turns to well who knew about it and did they share all that they knew? Were they being straight with the American public? That too suggest there's a logic or set of strategies or a set of political incentives that require explanation, that stand right behind the production of scandals.

Wioletta Dziuda:

Yeah and you think about this naive view of scandals. You would say when the scandal arrives, it signifies some misbehavior. So obviously people are going to hold the implicated politician accountable. We should see the support for them drops in the polls and so on and this is definitely not what we have seen consistently throughout Trump's presidency. We've seen a lot of scandals that actually did not really hurt Trump and if anything, it seems that sometimes he was actually benefiting from them. So we tried to understand when we think about the strategic incentives of political actors in creating scandals can we somehow learn something about in which circumstances those scandals will be very detrimental to the implicated politician in which circumstances they won't be.

Anthony Fowler:

I want to get into the details of the model, but before we do, you have an example of a scandal that you'd like us to keep in our minds as we read the paper, tell us about your canonical example that you want us to think about.

Wioletta Dziuda:

Yeah. We just have to be very careful because this is a family show. So we'll have to beep a few words.

Anthony Fowler:

Do we sensor, do we leave things on the show?

William Howell:

It's a good question. I was thinking when I was rereading our paper in preparation for our conversation here, I thought, "Wait, maybe this is the first time this word has ever been written in the American journal of political science in the history of the journal."

Anthony Fowler:

You're pioneers.

William Howell:

That's right.

Wioletta Dziuda:

So listeners, this is a trigger warning and with this warning, yes. So we'll go ahead.

William Howell:

So people may recall this. This happened in January of 2018.

Anthony Fowler:

Parents, if you're watching with children, you might want to mute.

William Howell:

Donald Trump gathered with some congressional delegates from both parties. Democrats, and Republicans to hold a discussion in which he was charged or accused of having disparaged immigrants from "shithole countries."

Tape:

According to a democratic aid familiar with the conversation. Mr. Trump was referring to African nations and Haiti.

William Howell:

This was meant to be a conversation about what to do in a particular policy domain and the meeting was private. It was not tape recorded, there is no definitive proof of what was actually said. But when people left that meeting, the Democrats promptly came out and said that Trump had talked about immigrants coming from "shithole countries."

Tape:

Democratic Senator, Dick Durbin was in the oval office when he says Donald Trump said things.

Tape:

Which were hate-filled, vial, and racist.

William Howell:

Whereas republicans said that they didn't hear it.

Tape:

Two other Republican senators who attended Tom Cotton of Arkansas, David Perdue of Georgia put out their own joint statement and in plain English took a pass.

Tape:

I didn't hear it and I suppose sitting no farther away from Donald Trump than Dick Durbin was.

William Howell:

The public then was left to try to figure out two things. First, did Trump in fact use this language. But then also who's lying? So there're these two unknown quantities that the public has to figure out and I think that becomes a helpful example for us to refer back to. Not because it's the most important scandal under the Trump presidency, but it captures the essential elements of the model that we eventually wrote down.

Wioletta Dziuda:

So what we are trying to do is we are trying to build the simple model that distills this anecdote and try to use this model to reason what the consequences of the party's behavior will be.

Anthony Fowler:

Let's talk a little bit more about the details of the model here for a second. So tell us about just the key mechanics of the model. They do pretty closely fit the shithole example, we'll call it.

Wioletta Dziuda:

Suppose there's a politician, let's call him or her president, but you can think of any politician. There are two parties. One party is aligned with the politician and the other part is misaligned. The starting point of our model are two observations. So first observation that we make is parties are privy to information about the conduct of this politician that the public is not. So when they observe misconduct, they have a choice to make, "Should we come out and say there was misconduct or should we stay silent?" When they don't observe misconduct, they also have a choice to make, "Should we make a fake accusation or should we just stay silent?" So for example during the meeting with Trump, if he actually did not use this foul language, Democrats still had the choice to make, "Should we just maybe accused him of disparaging those countries?"

It's believable. It might be believable. "Why shouldn't we try to do that? There's no way that voters can verify that we are lying." The second thing that we assume in the model that is that parties care about the reputation of the politician. So if you are the aligned party, you definitely would like this reputation to be as high as possible. If you had the opposing party, you actually would like to diminish the standing of this politician because then he or she will be less effective. The second thing that parties care about, here we can have a discussion. I think a lot of people feel uneasy by this assumption, but we think that parties would like to be perceived as parties with integrity and honesty. They would like to be perceived as parties who do not stand up for misbehavior, would tell the truth to the voter. You can understand why this is important. Voters would like to have parties that they can believe they will discipline their politicians, they will always tell us the truth what's happening. They would get rid of bad apples. Supporters want to keep reputation for telling the truth and that's basically it.

So we have those two parties that care about damaging or protecting the politicians. But at the same time, they care about protecting their reputation. They see what's happening. They see whether the politician misbehaves or not, and they make a decision, "Should we say something? Should we accuse the politician about the misbehavior or not?" And after they make their accusations or remain silence or defend the politician, voters observe what's happening and they make inferences both about the parties and the implicated politician.

Anthony Fowler:

So just to be clear, you guys have been talking about when you get scandal, when you don't get scandal. So just to be clear, when you talk about scandal, you're talking about scandal as being a report of misbehavior by either party. So when a party makes a report, it could have been that that was a fabrication. It could have been that in fact, there was no misbehavior, but a party made it up or conversely, if there was misbehavior, but there was no report of misbehavior by the parties, then that's not a scandal. So when you're talking about scandal you're talking about did one of the parties come out and accuse the politician of doing something bad.

William Howell:

This is a motivation for writing down the model as we've done, actually. It's not just a feature of the model, but it is a recognition that underlying corruption doesn't necessarily track trends in the production of scandals. Scandals are the public revelation of that misbehavior, but they mean to be one in the same and in the model we show the decidedly aren't one and the same. And that, to the extent that you use scandals as a rough proxy for how corrupt a politician is or a community is, you may be making really bad inferential errors. You see that when polarization increases, that leads to the production of more scandals, that then if you naively compare two cities or two politicians that are operating in very different political environments, vis-a-vis polarization, and you say, "Well, there's more scandals in one. So therefore there must be more misbehavior there." You'd be making the wrong inference.

Wioletta Dziuda:

Yeah and I think this is something that we weren't after trying to understand what are the negative consequences of polarization on our lives. But this is what we just got by trying to understand scandals. We found this novel prediction that polarization is not only bad because it feeds disagreement and unhappiness among voters, but it actually increases misconduct of the politicians. Because it voters think that Democrats are going to accuse Trump no matter what. Then if they see a partisan scandal, they say, "Well, it's probably the Democrats just accusing Trump. It's not the Republicans who are hiding information." That actually showed Republicans and that gives them more incentives to hide information. Okay, now the voters think that Republicans are hiding information. That is more incentives to Democrats to actually accuse Trump because when they accuse Trump voters will blamed the Republicans, not Democrats. So we found this actually quite interesting and depressing at the same time.

William Howell:

Yes. So when the stakes are higher, which is decidedly true during periods of polarization, what you can expect and what the model shows is that the number of scandals goes up, but it's a certain kind of scandal. So the number of what we call like bipartisan scandals, that is instances where both the aligned and the opposing party come forward and say, "This was misbehavior." That actually goes down, but the number of partisan scandals, where the opposition party says that there was misbehavior and the aligned party doesn't say anything at all, or denies the claim goes up markedly, and you can see why, right? Because the opposition party really wants to get whoever's in power out of power, or to undermine her effectiveness because the opposition party really disagrees with what's going or the purposes to which that power is being put. Meanwhile, the aligned party wants to preserve that person in power.

So you see an increase in the production of overall scandals, particularly as it relates to partisan scandals. Now, precisely because they're partisan scandals, you only got one party making the claim that there was misbehavior. The voter has a harder time inferring what actually happened. Now given all of that downstream, the politicians sitting there saying, "Huh, should I misbehave or not?" And so in the extension, we show that's precisely the instance where you should step forward and engage in this behavior because the voter is going to have a harder time putting the blame on you, that there's going to be claims from the opposition, but those claims aren't going to be especially damaging to you reputationally and so you have heightened incentives to engage in this behavior. So polarization leads to this cascade of really bad outcomes, heightened scandals, less information, more bad behavior.

Anthony Fowler:

Hey, if you're getting a lot out of the research that we discuss on this show, there's another university of Chicago podcast network show that you should check out. It's called Capital isn't. Capital isn't uses the latest economic thinking to zero in on the ways that capitalism is and more often than not, isn't working today. From the debate over how to distribute a vaccine to the morality of a wealth tax. Capital isn't clearly explains how capitalism can go wrong and what we can do about it. Listen to capital isn't, part of the university of Chicago podcast network.

Anthony Fowler:

So somehow scandals regarding the president, let's say are always hurting both parties, reputations in expectation. Why is that the case?

Wioletta Dziuda:

Well, when the voter sees parties disagreeing about what happened. The voter knows that one side is lying and because I don't know for a fact which one it is, I have to assign blame to both. Now I will differentially assign the blame depending on the circumstances. So for example, if I perceive the politician to be corrupt or incompetent, then probably I'm going to assign more blame to the party who is defending the politician. I'm going to believe it's more likely that actually the position indeed engaged in this behavior, but it's still possible that in this particular instance, this politician did not. It's the position party who is lying, so I'm going to lower my confidence in the honesty of both parties, at least a little bit.

William Howell:

Yeah. So then there also is another parameter that has to do with the likelihood that the parties learned about that misbehavior. So you might think about some actions as being highly obscure, and there's a very good chance that the two parties didn't actually learn anything about it. In those instances, observing a partisan scandal, the voter will look up and say, "Oh, well, there's a higher chance that this is just being fabricated. "So it's the oppositional party that's more likely to be dishonest. Whereas if we know that everybody was in the room, like with Trump's comments about the shithole countries. That's an instance where you look up at the Republicans, by all signs point towards them prevaricating about what they claim not to have heard. So it's both with regards to the underlying incidents of misbehavior and the probability that the parties learned about it, that you have differential fallout, reputational fallout in the aftermath of partisan scandals.

Wioletta Dziuda:

So what happens is we start Trumps presidency with some prior on how likely he is to misbehave. And then as the presidency unfolds, we keep on updating and if the presidency unfolded with no scandals, no accusations from either Democrats or Republicans, we would have concluded that he actually is a better president than what we thought. He's much cleaner, much better behaved, much more competent. If we on the other hand, so both part is accusing him over and over again, we would conclude well, he is actually very corrupt, very bad president and incompetent president. Now there's this middle ground where only one party accuses and the other party does not and what we find is that the effect of that is in between. So in this particular example, the aligned party really suffered the consequences of the scandal, but the implicated politician is sheltered. It might actually even benefit from that.

Anthony Fowler:

So just to summarize and be clear. So partisan scandals, anytime there is a partisan scandal where one party is saying, "Aha, they did something bad," and the other party is saying, "No, our politicians fine." That's bad for the parties reputations for the reasons that we've talked about, but that's not necessarily that bad and your saying could even be good in some respects for the politician. Why is it the case that Trump's reputation could actually benefit from Democrats accusing him of saying shithole countries?

William Howell:

So I think the first thing to note is that this finding that we're pointing towards can be recovered in an instance where your baseline probability of there having been misbehavior is really high. So in those instances where we're looking at Trump and you think, "Oh my God, this is precisely the kind of thing that he does all the time," that it may be beneficial. What the model shows is that there's negative reputational fallout for both parties, but the reputational fallout is greater for the aligned party than for the opposed party. Because the aligned party is looking up in this period of high polarization and says, "I'm standing by my president," and people expect the president to have engaged in this kind of misbehavior and the Republicans are sitting quietly by. As Dziuda points out though, that kernel of doubt, that's introduced by the Republicans unwillingness to come forward and to affirm the accusations by the democratic party, the opposed party, makes them think, "Huh, maybe in fact, this didn't occur." And that kernel of doubt leads to your reputational fallout, a little bit of reputational fallout for the democratic party.

William Howell:

That's what fuels the costs or informs the voters propensity to think, "Maybe they're making up this time," because everybody expects Trump to say this kind of thing and yet Republicans are saying no, so maybe this is that rare instance where he didn't misbehave. That thought that this may be that rare instance where he didn't misbehave, in voter's mind, it reduces their expectations or their assessments of the probability that he had in fact did misbehave.

Anthony Fowler:

Okay. So I guess one way of thinking about this is that there's essentially three different things that the voter could have seen. They could have seen both parties accusing Trump of misbehavior. They could've seen neither party accusing Trump of any misbehavior, or they could have seen what they saw in this particular instance, which was the out party accusing him of his behavior and his own party denying it. Essentially what you're saying is that if my prior beliefs are already pretty high that the president's engaging in misbehavior, then seeing that sort of middle category could actually slightly raise my beliefs about the honesty and the integrity of the president.

William Howell:

Integrity only has to do with the reputation of the parties

Anthony Fowler:

Right, because it slightly raised my beliefs about the chances that the president engaged in misbehavior.

William Howell:

Exactly.

Anthony Fowler:

Okay. So you think that's a pretty troubling state of affairs when we already have such bad prior beliefs about the president, that if only one party accuses him of misbehavior, then that could actually raise our belief. So Trump's still is harmed, trump's reputation is still harmed.

Wioletta Dziuda:

No, your assessment is going up. So what happens is we start with voters having some belief about Trump misbehaving, and now they will observe how parties behave and update their beliefs. If they see that no party is accusing Trump, they obviously have to conclude the probably Trump is better than what they thought. So Trump would benefit from those scandals, that's definitely true. Now what happens when we get a one accusation only from Democrats, so definitely this is worse for Trump that not facing any accusations, no doubt he would prefer not to have been accused. Yes. But voters when they see one accusation, they will have better perception of Trump that they had at the beginning of his term.

Anthony Fowler:

Yeah, okay.

Wioletta Dziuda:

There's three beliefs, no prior belief, the belief after a partisan scandal and the belief after no scandal. They are ordered this way, the prior belief, no scandal, partisan scandal.

Anthony Fowler:

The reason they're ordered that way is because in the unlikely scenario where there's a bipartisan scandal, then it goes down a lot.

Wioletta Dziuda:

Exactly.

Anthony Fowler:

But it's still the case that I think you answered my question incorrectly. If my question was, is it the case that Democrats accusing Trump of saying shithole countries benefits his reputation or decreases our beliefs that he engaged in misbehavior? The answer to that is surely no, it still hurts him relative to the counterfactual world in which they had not accused him of misbehavior.

William Howell:

But it benefits him relative to their prior assessment of the probability that he would misbehave.

Anthony Fowler:

Well, that's just an abuse of the word benefit. I mean, yes his beliefs are going up, but the only reason that our belief about Trump's reputation is going up is because the Republicans are saying the opposite. So it's both of those in concert with each other that are leading the beliefs to go up a little bit. But if I just wanted to isolate the effect of the Democrats behavior, the Democrats accusing him of misbehavior is clearly hurting his reputation.

William Howell:

Yeah. Well, there's two parts to this. You're pointing Anthony to the effect of the Democrats coming out and saying something. The effect on that is negative clearly. But it would be a mistake to assume from that, that that leads to an increase in their assessments of the probability of misbehavior. In fact, it's that it leads to a decrease in their assessments of the probability of misbehavior. It's just a smaller decrease relative to the case where in the Democrats say nothing at all and in that sense, the Democrats are able to preserve more of the public's skepticism and negative views about Trump. It's not about fortifying or increasing relative to the baseline of what they thought Trump was, increasing your assessments, "This is the kind of guy who misbehaves."

Wioletta Dziuda:

Yeah. I think one way to see it better is to think about Trump and his approval rating throughout his presidency. What our finding shows is that if at some point in time, we have this meeting in the white house and after the meeting, both parties come out and say nothing, don't accuse Trump of anything. Definitely Trump's reputation goes up. If Instead Democrats come out and say he disparaged those countries as actually they did then Trump is going to be worse off the situation, he's reputation will be lowered. It's still the case that any partisan scandal revealed that the particular point in time harms the reputation of the politician relative to what her reputation would be if no one has accused her of anything. But what our finding shows is that, if you look at the Trump's presidency, for example, over time and his approval rating over time, you might see that his presidency was plagued by partisan scandals.

Wioletta Dziuda:

Nevertheless, over time, his reputation was going up. The reason why it might have been going up, albeit maybe on the slightly is that because Republicans consistently have been defending him, that puts doubt in the minds of the voter that we know, we think Trump is misbehaving. Republicans wants preserve their reputation. Why aren't they jumping and trying to preserve their reputation by accusing Trump? So if we consistently Sera public has defending Trump, we have to be a little bit doubtful about how correct we were at the beginning of his presidency that he was actually this crook that we thought he was.

Anthony Fowler:

So I think you can say that scandals are bad for politicians, clearly. But in a counter-factual sense, a partisan scandal is not nearly as bad as a bipartisan scandal and it can even be the case that there are some circumstances in which, because our prior beliefs are such that we already think that Trump's likely to engage in this behavior. A partisan scandal might actually over time make it look like, oh the reputation of Trump is going up in light of a partisan scandal because it wasn't a bipartisan scandal. It could have been a bipartisan scandal and it wasn't. So if you were just looking observationally at data, you might naively infer that a scandal was actually good for the president, because look there was this partisan scandal and their reputation actually improved and that was because we thought it might be a bipartisan scandal, but it wasn't.

Wioletta Dziuda:

Yes. That's why we like this finding, because I think there was a lot of discussion in the media over time. Why is it the case that Trump is being hit with scandal after scandal? And when we look at his approval ratings, they don't seem to be going down, if anything, sometimes they were going up and people were coming forward with some explanations. Maybe Trump is just a different kind of person, maybe the normal rules don't apply. What our paper shows is it's not, that actually this is something that might happen.

Anthony Fowler:

And there's another interpretation of this, which is having observed that. So having observed this thing where Trump has scandal after scandal. His approval rating doesn't seem to be dropping very much. We could say, "Okay, well. I mean, surely his approval rating would have probably been increasing if he hadn't had all these scandals." So in a counter-factual sense, he's still suffering from these scandals.

Wioletta Dziuda:

Yes.

Anthony Fowler:

And furthermore, there's probably another by-product of this, which is that the reputation of the Republican party is probably being harmed as all of these scandals are happening because they're continuing to side with him throughout more and more scandals, that leaves the voter not sure whether or not Trump is really dishonest or not. But as more and more of these scandals keep popping up, every time one pops up they say, "Maybe the Republican party is just a completely dishonest party that's willing to lie for Trump." So it's not like the flat approval rating for Trump is a sign that all is well, and the scandals are having no effect. It could be as he's essentially causing damage to his party the whole time and he certainly hurting himself relative to what he could have achieved without all this misbehavior.

William Howell:

Yes, in the case that we're describing here, the entity that suffers most from a partisan scandal is the aligned party. When you think about this iterating over time, the notion is this can't possibly be the case that in a world in which I strongly expect the politician to be misbehaving and the Democrats in this case again and again, are saying as much that the Republican party again and again is insisting otherwise. It has to be the case that they're then the strategic type, like they're revealing themselves as such. In that sense, in the context of the model, the reputation is all about one's ability to appear honest and so they experience all the fallout.

Anthony Fowler:

An obvious question and one that your paper answers I think pretty well, is why would they do this? I think this is a question lots of observers have been asking about the Republican party throughout the Trump presidency. Why are they letting their reputation suffer by siding with Trump over and over again?

William Howell:

Because they also care about the politician and the politician remaining in power. So if the alternative is Trump not being able to carry forward and use the powers available to him in order to affect change, then that too comes at a terrific cost. That's what the parties are weighing. On the one hand, they would like to be perceived as being honest, but they also get benefits from being aligned with a politician who's in a position of power and can affect meaningful change.

Anthony Fowler:

But there was a period where they could have had Pence instead of Trump. If we want to pontificate for just a second. There was a period where the senators could have voted to convict the president and they would have had Pence as president instead. That seems good for them reputationally in terms of leading voters to think, "Oh, maybe this is a more honest party with integrity." Also the cost from a policy standpoint is not very high. Maybe it's even a win for a lot of Republicans. Why did they not do that?

William Howell:

Yeah. I think there's this other piece, which is right now, the reputations of the parties are strictly about the reputations for honesty. But you could also think about them having a reputation for what you might call judgment. Their willingness to get behind and align themselves with honorable folk and the natural thing that the Republican party would face if they came forward and said, "Yes, in fact, Trump is a scoundrel," is the question, "Well then why did you ever bring him on board? Why did you ever align yourself with him?" Now that's outside of the model, those kinds of considerations are outside of the model. But it's another kind of reputational concern, which then would bear upon their ability to win in the next election and then for reasons that are within the model, we can expect there to be this big shift in who holds the office which is then going to drive them to nonetheless stick by Trump, even though we have available to us like Pence and everybody sees that his baseline propensity for misbehavior being much lower.

Wioletta Dziuda:

Yeah. I think that's right and I think this is something that Will and I are thinking about trying to run a formality later on. But you can also think about the Republicans not thinking about who is the president currently. But what are the chances of Trump and Pence respectively to be re-elected. So if they actually impeached Trump and if they convicted him and got Pence, perhaps they knew or they thought that come next election, Pence had much higher chances of losing because of just who he is and how much support he has than Trump. So they were thinking not about the current benefits from having Trump in office, the Republicans in office, but they were thinking about the benefits that they can get after the next election and then those benefits are high because it's their president versus Bernie Sanders in their minds and then the story goes through.

Anthony Fowler:

All right. So you guys don't get a bottom line for this paper, but I'll give you.

Wioletta Dziuda:

No, we want to have the lighting round.

Anthony Fowler:

You get one minute to say something about your paper that we didn't discuss before. We won't call it a bottom line, we'll just call it an author's last plea to the listener. Anything else you'd like to convey about your paper before I turn it to the discussion part?

William Howell:

I guess I would say two things and these lessons are at the margins of what we've been discussing, but it's worth bringing them forward, front and center. The first being that the strategic incentives of parties to be truthful or not truthful depend upon the specific structure of the informational environment and the nature of the scandal that we're talking about. So in the paper, we spent a lot of time trying to be clear about the class of scandals that we have in mind. One can imagine different kinds of scandals where the implicated parties have different relationships with the politicians that are going to lead to players to behave in different kinds of ways. In that sense, and this goes to the second point, generalizing about scandals and certainly as we've said, trying to make inferences about trends in the production of scandals. Inferences about those kinds of trends about underlying rates of misbehavior is just a really fraught enterprise, precisely because of all these incentives that that politicians face. So the point of the model is to try to think really clearly about the politics of scandal production, and to warn the reader, the observer, to be careful about making strong inferences about misbehavior, just on the basis of how many scandals she or he observes.

Wioletta Dziuda:

And I want to go back to one thing that you Anthony pointed out throughout this interview. You push us a little bit and you said you can't explain everything. There are so many instances where it doesn't seem that your model really gives us a very clear idea of what happened. I think it's worth emphasizing that we are completely aware of that, but what we want to convey is that if you start thinking about a scandal as something that arises endogenously, that's generated strategic actors, even in like a very simple setting, you already get a lot of nuanced filings that actually explain better what we see outside than this naive approach. So I think my hope would be not to just now sit down and say, "We've explained everything, and you can now take our paper and try to understand what's happening," but it's to encourage people to think in a little bit more systematic way about scandals, because I think there's a lot that we can gain from that.

Anthony Fowler:

All right. that was it. Now that Dzuida and Will are gone I can have the discussion all by myself. I think it's a really interesting paper and I think it is actually, even though it's a simple model as you say, I think it has a ton of applicability to lots of real world political scandals and it allows us to think about lots of very interesting political trade-offs and political strategic decisions that are made all the time. So I think this model and this kind of thinking will probably be very useful. I don't know exactly what implications it will have yet, but I think people should take this data and say, "Okay, what can we say," given that we don't get to observe the misbehavior, but we do get to observe the scandal and we do get to observe when the scandals are essentially supported by just the out party or both parties. I think that's a really interesting thing to think about when we should expect these things to arise. What kinds of things should we observe in the data if we believe the mechanics of this model and so forth.

So I think this is a really interesting paper that does actually tell us a lot about modern American politics and why it might seem like we get political parties that behave in such a partisan way and why also politicians seem to not suffer very much from all of these scandals. So I think this helps me understand that in a deeper way than I did before. So thank you guys for doing this. This is fun.

Wioletta Dziuda:

Thank you.

William Howell:

Thanks for putting up with us and leading us through the paper. We are so broke.

William Howell:

Thanks for listening to not another politics podcast.

Wioletta Dziuda:

Our show is a podcast from the Harris school of public policy, and it's produced by Matt Holdup. Thanks for listening.