Theory of Change Podcast With Matthew Sheffield Democrats are on an election win streak despite having a badly damaged brand, what’s going on?
Donald Trump was swept into office by an elaborate series of lies about his radical policies, but more than a year into his second term, the less-engaged independent voters who powered his victory have turned firmly against the president.
But as low as Trump’s approval ratings have fallen, the Democratic Party’s favorability among Americans is even lower.
How is this possible and what does it mean? Depending on who you ask, you’ll get a very different answer. Usually, however, the criticism boils down to: Democrats aren’t promoting my own personal policy opinions.
The hard truth, however, people don’t want to accept is that many, if not most, voters have policy viewpoints that aren’t fixed, which means that focusing your campaign strategies based solely on public opinion is not going to work.
Democracy in America is severely endangered because one the country’s two major parties has become a fascistic personality cult. But a strategy of protecting democracy by winning every election forever is doomed to failure.
So what to do instead? That’s an answer that I can’t give you in a single podcast episode, although be sure to subscribe nonetheless! But what I can say is that democracy defenders must think bigger and be much more open to new voices and new ideas.
And joining me for today’s conversation is a friend of the show, David Atkins. He’s a member of the Democratic National Committee and also a contributor to Washington Monthly.
The full discussion of this episode is for paid subscribers. An excerpt on YouTube is available, but you will need to be a premium member on Patreon or Substack to watch, read, or listen to the full discussion.
You can subscribe to Theory of Change and other Flux podcasts on Apple Podcasts, Spotify, Amazon Podcasts, YouTube, Patreon, Substack, and elsewhere. (Note: Purchasing a book through the links in show notes helps support Theory of Change.)
Related Content
--The 2025 elections showed that more than anything, people want Democrats who fight Trump
--Republican operatives completely reconfigured politics, their Democratic rivals have not kept up
--How the American left became post-political, and how to change it
--Republicans built a massive infrastructure to attack democracy, Democrats have not made one to defend it
--Democrats get lots of bad advice, particularly the idea that most voters are ideological
--In 2024, Donald Trump bet big on ‘unlikely voters’ who have sat on the margins of American politics
Audio Chapters
00:00 — Introduction
06:13 — QAnon as a religion of narcissism
12:18 — What conspiracism offers middle-aged and older women
20:13 — Media proliferation and political manipulations have made conspiracy belief much easier
28:27 — The women of January 6th faced widely divergent economic circumstances
34:32 — Charismatic evangelicalism as the common starting point for QAnon believers
44:02 — Astrology, space aliens, and QAnon
48:44 — ‘Soul contracts’ and tragic morality
52:49 — Right-wing politicians harm society and then use the nihilism they engender as campaign leverage
55:42 — What do QAnon believers think about the Epstein files now?
01:04:35 — Prevention is easier than de-radicalization
Audio Transcript
The following is a machine-generated transcript of the audio that has not been proofed. It is provided for convenience purposes only.
MATTHEW SHEFFIELD: And joining me now is David Atkins. Hey, Dave. Good to see you back on the show.
DAVID ATKINS: Hey, happy to be here. Thanks for having me on.
SHEFFIELD: Yeah. So you are a member of the Democratic National Committee, but you are here in your personal capacity.
ATKINS: Yes.
SHEFFIELD: So we want to make sure to point that out.
ATKINS: Yes. I’m not an official spokesperson for the DNC in this interview.
SHEFFIELD: Yes. All right, so with that out of the way one of the topics that I wanted to talk about here today is that Donald Trump, as I think everybody by now, or at least people who watch or listen to this podcast, knows and read you is that Donald Trump, he won not because of the fanatical fascistic, right wing, he won because he kind of misled a lot of people who didn’t know very much about politics, and those marginal Trump voters appear to have pretty much turned against him at this point.
And his approval ratings are the lowest they’ve ever been. And in some polls actually even lower than they were after January 6th 2021. So, But the paradoxical thing is that the Democratic Party is rated as less popular in polls.
And I think that’s, it’s causing some people to kind of project, a lot of their own personal biases onto that data set. But there’s a lot going on there. And, but ultimately, I mean, voters are still, they’re still choosing the Democrats in elections.
ATKINS: Right. I think it, There’s a lot on there. When you look at a statistic like such and such number of people dislike or like the Republican party or such and such or like the [00:04:00] Democratic Party that is genuine, that is generally. A confluence many different Factors. there’s an old sort of in, in religious studies that every religion is sort of like a flashlight on the elephant, that everybody’s sort of got a spotlight that nobody can see.
The whole elephant. I think you have a similar thing going on here. So you have the moderates who are saying, oh, this means that the Democratic Party is too far to the left and need to come back to the center. you’ve got leftists who are saying, well, the Party is bad on this issue or that issue.
And if they were only farther left, I think it really depends the person. I think it’s all of those things are true for different segments of the electorate, which makes solving the problem challenging. But I think one thing you can say is there are a few major reasons for happening. Number one, you have a low trust society in general, so all institutions are suffering across the board. Approval of every major institution is down.
That having been said, not making excuses for the state of the, Democratic Party approval. ‘cause I’ve been talking to various leaders in the party about a lot with some alarm, I think. Yes, right. There some people who have joined the Trump Coalition who used to vote for Democrats, who feel that the party has shifted too far left on issues.
But rather than take the Yglesias sort of angle this that has been happening since the 1960s, you have been having realignment shifts this for the last past 50 or 60 years. And that doesn’t mean that you need to stop expanding rights or do or stop advancing social change. And in any case, it’s not Democratic candidates or the Democratic Party officially that is advancing civil rights in this way, that is making those voters uncomfortable.
So there’s only so much the party per se can do about that. So when Matt Yglesias and those folks say, oh, the party needs to shift to the right, I mean. They’re not talking about party candidates, they’re talking about random [00:06:00] activists on social media. So good luck, I guess. There is also another segment of people who are absolutely furious Gaza or some other issue.
And again, though, know, you can’t really fault candidates so much for this, and candidates who have taken much more left positions on those issues are not actually fairing better in elections by and large, with some exceptions. And we can talk about the Mamdani Coalition and all of that, and I’m very supportive of a OC and Mamdani and those folks.
But it’s not exactly an electoral panacea. It’s not like if every candidate adopted those positions, the party’s fortunes would be reversed. It’s not that simple either. I think the biggest thing that is impacting though approval of the Democratic Party, ironically, is from core normie Democrats. You ask a core Normie Democrat who shows up to a No Kings protest.
And is with Trump, if they approve of the Democratic Party, by and large, they’re going to say no. Not because they like Trump, not because they think the party is too far left or too far right, but because the party is not doing a big, a good enough job of standing up to the Republicans. And look what some strategists and Chuck Schumer might say is, oh, we’re doing exactly what we need to do to win elections.
Look how well we’re doing winning elections. Well, okay, maybe, but there’s more to politics, ironically than winning the next election, right? You’ve gotta keep people engaged and believing in you as an institution, believing in your values. Otherwise, you’re just going to get a thermostatic effect where, okay, you win the next election, but the next time people get upset over inflation or whatever, you lose again.
And if your entire premise of how you defeat fascism is we have to win every single election. Rather than we have to end fascism at its root, then you’re going to lose. So there has to be more than [00:08:00] just, oh, we’re doing whatever it takes to win the next election by looking calm and looking like good guys.
So it’s a lot of things I think.
For voters, ideology matters less than activity
SHEFFIELD: Yeah. Yeah. And we’ll come back to the thermostatic issue later. But yeah, I, it’s it is shaping up when we look at the candidates who did win in the past 2025 election the major candidates who won, what we’re seeing is that, yeah, that the real access of approval for Democrats or energy is yeah, how is, how much you want to oppose Trump.
And it’s not as even as much of an ideological barrier. So like we see, for instance, with Virginia’s Abigail Spanberger has, recently come out with some pretty tough restrictions on the Trump Ice Thugs and what they’re allowed to do legally within the state. And, the degree to which Virginia law enforcement officers are allowed to cooperate with them or provide them information.
And, so this is somebody who in the conventional left right intra Democratic Party splits. This is not somebody who is on the further left of the party, but on the other hand, she also shares that desire to vigorously oppose the authoritarianism of Donald Trump that, that Zoran Mamdani does.
ATKINS: Right. No, exactly. And this is one of the things, like I, I was not a big span Spanberger supporter because she was on sort of the moderate side of a lot of policy fights that I was not approving of. But look at what I mean. Now I’m a big span Spanberger fan because hey, like those, she’s, that she’s not annoying me on any policy fights in Virginia, but what she is doing is standing up really strongly to Trump and ICE.
And I couldn’t be happier about that. And I know a lot of other folks who were span Spanberger skeptics who are very happy with her as well. And think that if we have more of that in a real way, I think, that will also be helpful. [00:10:00] You do see a lot of politicians sometimes in a cringey way coming out and using, F-bombs this kind of language and, using stronger language now, which is nice to see, as long as it actually feels natural.
But you know, actually stepping forward and. Demanding say to visit ICE detention centers or actually, stepping forward and throwing real sand in the gears of the Trump regime. what people are looking for.
SHEFFIELD: Yeah. And it, and to kind of boil it down a bit here, what we’re seeing is that there’s kind of a, I would say that, there, there are your policy views, there is your operation style and then there’s your communication style. And those are the three things that, that people are really caring about.
And, and what it’s looking like is that there is kind of a, a real alignment that’s shaping up in terms of communication style and operation style, that people are realizing. The bigger problem here is that we have to stop fascism first, and then build the case simultaneously for, a society that, that does address the issues that people are are concerned about and, but also is willing to talk about democracy. because like, I guess that has been a debate point as well within the party that a lot of people have said, well, the public doesn’t care about protecting democracy. And other people say, well, no, they do.
And it’s, I, it, I don’t think you can say one way or the other. It’s a matter of how you do it, is what I would say.
ATKINS: No, I agree. And I think that G Elliot Morris, for instance, has had some very compelling data recently that people do care a great deal about protecting democracy. It also really matters. And this is sort of a, a cart/horse like [00:12:00] chicken/egg egg kind problem. In the sense that if you take the popularist view, which is based in large part upon a bunch of quantitative survey data, and we could go into all of the challenges with quant data.
I, I’m a qualitative, research guy by trade and man, like the mistakes that you can make just by paying attention to what a quantitative survey says are enormous. But of course, if you ask people on a quantitative survey what they care more about the price of groceries or, threats to democracy, most people are of course going to save the price of groceries.
But there’s a huge emotional investment in democracy as an idea. And if your leaders are not talking full throated about the problems and the threats to democracy in a way that sounds more like, that, sounds like more than just. The heated political rhetoric of the day. But if you manage to show people no, like you are actually not going to be able to vote for your leaders, you’re not, there’s actually going to be an accountability problem in the, in your democracy.
And these people are trying to set themselves up to rule for life. People do care about that. People do want to step to defend that. And we and what’s been shown in the data is not only are people actually concerned about this and increasingly concerned about this more than they were six months ago, in part because of the actions of the Trump administration, but also because when you have leaders not named Chuck Schumer, but actual thought leaders who are now actually more credible on the left and within the Democratic Party who are actively talking about this, people pay attention.
Journalists pay attention, it becomes more part of the conversation. And lo and behold, voila, people start to care about it, even in the quantitative survey data. So you don’t just have to reflect whatever the public opinion is from six months ago. In a survey, you [00:14:00] also have a role in talking about the issues of the day and shifting public opinion because you’re not, and not even in a way that changes people’s minds, but that changes the salience of the issue.
That changes their focus and their understanding.
SHEFFIELD: Yeah, absolutely. And we’ll get into that, but I want to circle back to the problem with claiming that your ideas are just pure math, which is what a lot of people that especially of the the self-identified popularist that’s they often say that they’re just doing math but they’re really not.
And but even aside from, the fact that they are trying to promote their own ideological preferences, which they never state that, but setting their preferences aside though, just the, and I can say this as somebody who, used to do polls and write about them.
So obviously I, I think polls are very useful and important. But they’re far less scientific than people imagine them to be, in part because just the very act of taking a survey is altering your mindset. And so it’s take, it’s taken you out of your regular mindset of your, which it would be your ballot voting mindset.
But it’s also, it is a, it requires the pollster and the person to have the same understanding of the question, and there’s no proof that is true.
ATKINS: No.
SHEFFIELD: It’s like it relies on a fundamentally qualitative assumption, without ever saying, saying it.
ATKINS: Right. Yeah. Look, I mean, people have complicated thoughts about politics generally, especially undecided voters or cross pressured voters, people whose vote is not already taken for granted. Typically, your partisans, I mean, they also have your partisans, your deep partisans also have complicated thoughts about politics, [00:16:00] but also somewhat more predictable, but especially people whose votes are winnable on either side.
Tend to be either more ignorant of the issues or they tend to be really crushed, pressured and conflicted, or both right to where if you ask them to explain themselves, if you ask them to explain what they think about an issue, you’ll get some confused and maybe some contradictory, but also some, complicated views on subjects like education or maybe, trans rights or maybe, taxation or what, or housing or what have you.
When you boil that down into a response to a poll question, right, and you’re a sophisticated pollster, you understand this, first of all even just baseline, the way you ask a poll question can have enormous biasing effects into the answers that you get. get. But even at that, this sort of goes back to George Lakoff and frames of the world and all that, people operate with a lot of different frames in their mind about how the world works.
And depending on which frame of how the world works, you’re activating, sometimes contradictory, sometimes parallel. People can come to different conclusions about what is important or how they want to perceive the issue. That you cannot possibly reflect in a bubble answer on a quantitative poll.
And you can get those to say almost anything that you want within reason. Whereas if you ask someone, Hey, what do you think about housing? What should we do about housing? In your ideal world in a focus group, that you’ll get a lot more honest answers. Of course, then you are, subject to the interpretations of a focus group when a consultant decides to write a report, but by and large, you’re going to get much better idea of [00:18:00] how the world works and how the electorate functions by just listening to a cross section of maybe 60 undecided voters than you will getting the captured survey responses from a thousand.
It may be statistically significant per the mathematics of stat of stats, but the gar, but the data you’re getting is garbage. It’s in, garbage out. For the most part. It, well, it’s not total garbage, but it’s not nearly as useful as a guideline as people want to believe it’s.
The limits of polling and quantitative data
SHEFFIELD: Even if they did understand the question you might be catching them in a moment where one particular opinion of theirs about this issue is more salient in their mind. And then if you were to talk to them the next day, another aspect might be more salient depending on whatever their circumstances are.
And so. And to be fair, polls do always say that this is just a snapshot in time. And I would say that the actual polls themselves are far more nuanced about what can be learned from polling than the popularists who kind of have like a, I mean, I call it cargo cult social science.
Like that’s what they’re kind of doing. They’re not, most of them do not do polls themselves with some exception. And so they don’t under, they don’t have direct experience at how fungible everything is, even though they know in principle about question wording distortions and whatnot, until you’ve actually seen it with your own I just, it I don’t see, I don’t see it as critical.
ATKINS: Right. And I think that the last thing that’s really important, and that’s become very obvious this last year of the Trump administration, is there’s a very big difference between talking about theoretical policies on paper. And the actual implementations of those policies when it comes to [00:20:00] their real and emotional impacts.
Right? So say immigration, it’s one thing to ask people when the general media environment on social media, Fox News, and everywhere else has been ramping up this mass hysteria about, about, immigration to say, oh, do you want to deport all people who are not here legally and you’ll get a high number?
Do you want to close the border entirely and deport everyone who lives here? And, in advance of the 2024 election, you were seeing fairly high numbers for that, which led people to say, oh, the Democrats need to move left on immigration. But the problem is the actual implementation of that policy is horrific, economically destructive, socially devastating.
Nobody what people have seen in the attempt to implement. Even A part of that policy, they hate what they’re seeing. You do the same thing for trans rights, right? Like where you, where ultimately you have to go down to what, genital inspections of teenagers like you, you’re the actual implementation of policies that might sound good to people on paper end up being horrific in practice.
And it’s one thing to ask about that in theory. It’s another thing when it comes to the real world of politics and whether you want to allow those dry policies on paper and the questionnaire responses of about that to drive the way you talk about it in a debate or on a policy stump speech or in an advertisement where you explain what these policies actually mean.
You can’t be scared by a 54% approval number for a horrific policy. You can’t be scared about talking about what that actually means in terms of implementation. The Democrats did a terrible job of talking about what these policies would actually mean, which meant that Stephen Miller and his people thought they had a green light to do horrific things.
The public doesn’t actually like
The thermostatic nature of public opinion and Republican deception
SHEFFIELD: Yeah. [00:22:00] Well, and that does, go to the thermostatic nature of public opinion. So, so within pub political science just for people who are not familiar with the term, thermo, the theory of thermostatic public opinion is that. A lot of voters, perhaps even most are motivated more about opposing things that they don’t like, than than having an affirmative vision.
And so there is this kind of core, large core of voters who are persuadable by either party. Because whenever the par, whenever a party gets into power, they do things that, can be, the, these are the actual instantiations of the ideas. And sometimes people are like, oh gosh, I didn’t want that.
And so, and we’re seeing a lot of that. Yeah. As you noted with regard to Donald Trump, that a lot of people are saying, well, I didn’t vote for him to do this. I didn’t vote for him to, cut cancer funding. I didn’t vote to, to ban federal funding for vaccines. I I didn’t vote, so they’re saying I didn’t vote for that.
But in reality they did. They just were not educated enough about the positions of Trump on these issues. And that the thermostatic nature of public opinion, I think is, has, it has been a problem for Democrats because for Republicans are so deceptive and willing to lie about their policies and they’ve always been, since, ever since Mary Goldwater got wiped out in 1964, they’ve kind of realized, oh, well we can’t be upfront about what we actually want, and so we’re just going to, speak in generalities.
Vague terms about people being responsible and and law and order, and stop talking about their actual full positions.