Great piece! And, oh man, the whole paragraph around this: "They’re people whose social identities, professional networks, and moral self-understanding are organized around being the kind of person who cares about climate, who stands on the right side of history, who supports the arc of progress." And, I think one could do some really cool political science on that insight that nicely expands existing social identity frameworks.
The identity-protection mechanism is doing a lot of work in the piece and I think it's genuinely undertheorized in the political science literature. We have good frameworks for partisan identity (Mason, Huddy), for elite cue-taking (Zaller), for motivated reasoning (Kunda, Kahan). But the specific dynamic where professional epistemic identity becomes the thing being protected, where what's at stake isn't "my party" but "my status as a serious, informed person," that's a different animal. It operates upstream of partisanship. You can see it in how academics sort which findings are "interesting" versus "problematic" before any political content enters the picture.
Someone should be writing the paper that maps how knowledge-class professional identity interacts with information evaluation in ways that aren't reducible to partisanship or ideology. There's probably a survey experiment in there: give people identical empirical claims sourced from high-status versus low-status institutions and measure not just belief updating but willingness to engage. I'd bet the identity threat shows up as avoidance rather than disagreement. People would say "that's not serious."
If you write it I want an acknowledgment footnote.
There's plenty of people who criticize the Democratic Party and want it to moderate who are quite successful - Matt Yglesias for example. Hell, Andrew Sullivan is a Tory who continually complains about Democrats but is doing quite well. To point at a new-ish publication, The Argument has many of the same criticisms of the Democratic Party, but seems to be doing much better.
The difference is they actually talk about when Democrats, even leftie Democrats do good things and yes, also point out when the Republicans do terrible things. You need some honey to get the medicine to go down.
The Liberal Patriot had one story - the Democrats are too woke and left-wing and they're going to lose because of this, and it turned out, even among people upset with the Democratic Party's current trajectory, there was a limited audience for a 100% anti-Democratic Party tone and you could see that in the comment sections, which were only trolls or people obsessed with anti-woke viewpoints, which is just as outside of the median Democrat as some left-wing bluehair in NYC.
I'm a dirty leftie so I was never going to agree w/ them, but there were multiple people who were moderates who pointed out the issues w/ the TLP model. Ruy as a voice in a broader center-left/centrist publication might've worked, but in shocking news, if your only story is "Democrats suck and need to do better," when even in Texas, people like Kamala have +74 approval ratings among Democrats, there's not much of a constituency outside of the very small amount of obsessive anti-woke centrists, who are disproportionately on sites like this and Twitter but are frankly out of touch with the median Evelyn Normielib as lefties may be out of touch w/ the median voter.
This is a fair critique of TLP as a publication and I don't disagree with most of it. The tone did become pretty one-note, the comment sections were rough, and you're right that @Matthew Yglesias and @The Argument do similar work with more range and more success. If I were writing a media analysis of why TLP failed as a business, your explanation would be most of the story.
But that's not quite the argument I'm making. The piece is about what happened before TLP existed — when Teixeira was a senior fellow at CAP, Brookings, and AEI, publishing data-driven analyses of coalition realignment. The donor defunding and institutional exit preceded the Substack, and the Substack's tone probably reflects someone who'd already been pushed out of the spaces where nuance gets rewarded. That's not an excuse for the one-note quality — it's a sequencing point. The publication you're describing was partly a product of the pattern I'm describing.
On the constituency question — I'd actually agree that pure "Democrats are too woke" content has a limited audience, and that audience skews weird in exactly the way you describe. But Teixeira's early work wasn't that. The Emerging Democratic Majority was the most optimistic Democratic coalition analysis of its era. The shift from that to what TLP became is itself part of the story.
The comparison to Yglesias is interesting because it kind of proves both our points. Matt threads the needle — he'll criticize Democrats on process and strategy while clearly being on the team. He's also independently wealthy and doesn't depend on institutional funding. That independence matters more than tone, I think.
I think the contrast between Yglesias, the Argument, and TLP is really interesting. I love MY and pay him money and so far am holding out on the Agument because I already have too many subscriptions, but also love it. And, I often read TLP and appreciated it, but never to the same degree. And I think it's because TLP needed to expand beyond "Democrats suck" in ways that Yglesias and Demsas consistently do. Also, it's not like MY is operating off a trust fund, this needle he threads is lucrative. There's a market for smart, centrist criticism that gets that Dems on trans issues are dumb, but Trump on almost everything is 10x worse.
I think the market comparison is actually revealing in a way that supports both the above point and mine. Yglesias and Demsas succeed because they do the full range. They'll tell you Democrats are wrong on housing and then spend the next post explaining why the Republican alternative is worse. TLP didn't do that, and the publication suffered for it. No argument there.
But I think the interesting question is why TLP didn't do that. Some of it is just editorial choice and temperament. But some of it, I think, is what happens to your analytical range after you've spent years being told your empirical observations are unwelcome. Yglesias built his platform from a position of institutional security and social capital. Teixeira built his after being pushed out of the institutions where that range would have been natural. The Substack was the product of the pattern, not just a parallel case.
The market point is real though. There clearly is a big audience for smart center-left criticism that maintains team membership. The fact that Yglesias can be as heterodox as he is while staying legible as "one of us" is partly skill and partly the financial independence to not care if a donor flinches. That second part is underrated in these comparisons.
So here was the mystery to me.. of all the pieces Teixera wrote about various D aligned issues- why was climate the straw that broke the proverbial camel’s back. The effective policy “Sine qua non”? Could it be that “climate” and its only D-preferred solutions somehow are fundamental in a way that we are unable to discern? If we go back to the Hartwell paper in 2009, there have been a variety of solutions and priorities to decarbonizing, but only a narrow range are acceptable to the donor group. We could ask them “why?” except donors are not accountable to citizens.
The Hartwell Paper reference is sharp, and it's worth unpacking for people who don't know it. That paper (2010, out of LSE and Oxford, authored by 14 climate and social scientists including Pielke, Hulme, and Rayner) argued that the Kyoto framework had "dangerously narrowed the option space" for realistic climate policy. Their core claim was that treating climate as a discrete problem to be "solved" through one pathway (emissions caps, rapid fossil fuel phaseout, net zero targets) was both scientifically reductive and politically suicidal. They proposed multiple parallel approaches — low carbon taxes funding R&D, tackling black carbon and tropospheric ozone, forest protection, universal energy access — on the theory that decarbonization would only succeed as a contingent benefit of goals that were actually politically attractive, not as a moral crusade demanding sacrifice. (Tagging @rogerpielkejr, maybe he'll have something to say about this...).
So, yes, that's Phase 1 thinking applied to climate. Empirically grounded, pragmatic, treating the problem as persistent and complex rather than as a test of moral seriousness. And the Democratic donor class's response to it over the last 15 years has been almost exactly what you'd predict: the narrow pathway became orthodoxy, the Hartwell alternatives got filed under "not serious," and the range of permissible climate positions contracted until even TLP's relatively modest deviations were enough to trigger defunding.
Your question about why climate specifically was the tripwire is the one the piece probably underexplores. My read is that climate is where identity and policy are most tightly fused for this donor class. Immigration is a position you hold. Climate is a vocation. It organizes their philanthropy, their social networks, their sense of moral purpose. The other issues could be debated because they were held at arm's length. Climate couldn't be debated because it was held close. Questioning it didn't feel like a policy disagreement. It felt like an accusation.
And you're right that donors aren't accountable to citizens. That's part of what makes the Shadow Party so effective as a Phase 3 enforcement mechanism — it operates outside democratic feedback loops entirely.
First, let me state how much I'm enjoying your newsletter, which I found through someone linking to your "Reality Bats Last" post; you should have more than a few hundred subscribers!
As a Democrat for 50 years now, I'm sympathetic to your diagnosis, and find it interesting to contemplate how it applies to other issues, most notably education and crime. In the area of education I'm interested in reforms in states like Alabama and Louisiana that are apparently successful in helping low-SES students, as well as new tech-enabled education models like Alpha School, Math Academy, etc. However I suspect the chances of either having an effect on public education in my own state of Maryland are slim to none, given that they threaten both the prevalent orthodoxy on how public education should work and the economic interests of a major part of the Democratic coalition, namely public school teachers and related education professionals.
On the other hand, it's interesting to see Democratic officials in heavily-Democratic cities like San Francisco pursue policies to crack down on crime and public disorder, an issue that's traditionally GOP-coded. Perhaps that's because the costs of crime and disorder for that part of the Democratic coalition consisting of city-dwelling professionals are perceived to be high enough to outweigh the influence of other parts of the coalition that have advocated for restorative justice, prison abolition, defunding the police, and similar policies.
Thank you — that's very kind, and I'm glad "Reality Bats Last" brought you here!
Your two examples are doing exactly what I hoped the piece would prompt — applying the Phase 3 lens to other policy domains and seeing where it clarifies things.
The education case is a great illustration of Phase 3 operating at full strength. You've got empirical results from Alabama and Louisiana that would ordinarily interest anyone who cares about low-SES student outcomes — which is supposedly the whole point. But the results come wrapped in policy frameworks that threaten both the orthodoxy (how learning should be organized) and the coalition (who gets to organize it). So the evaluative question quietly shifts from "is this working?" to "does this empower the right institutions?" That's the utility-evaluation move I describe in Part 2. The evidence doesn't get refuted. It gets filed under the wrong heading and ignored.
The San Francisco crime case is interesting because it might actually be the exception that clarifies the rule. When the costs of a policy position land directly on the coalition members who hold institutional power — city-dwelling professionals dealing with visible disorder in their own neighborhoods — the feedback loop still works. The signal gets through because the people receiving it are the same people who set the agenda. That's not Phase 3 breaking down. That's Phase 3 never fully forming on that issue, because the constituency bearing the costs had enough institutional leverage to keep the empirical question live.
Which raises an uncomfortable implication: the coalition is most responsive to evidence when the professional class is the one paying the price. When the costs land on working-class members (lost energy jobs, rising insurance premiums from climate policy, school systems that don't work) the Phase 3 filter holds. That's not a conspiracy. It's just how social sorting interacts with institutional power.
Being a Canadian makes me somewhat hesitant to comment, but your question "what did we do that made him look like the better option?" made me stop and think.
First, it can be equally applied to the Republican Party as well. What did they do to allow Trump to take over the party?
From the outside, it seemed that Trump told the voters he understood they were unhappy and he would bring real change.
I think the lack of change in the Democratic Party is the answer to the original question. Hence, why their refusal to change is probably the most dangerous thing they can do.
And you're right that the question applies to the GOP too. The Republican establishment's failure to take working-class economic grievances seriously created the vacuum Trump filled. The difference is that Republicans adapted (badly, in ways that traded principle for power) while Democrats are still debating whether the grievances are real. That's the asymmetry that worries me.
Your instinct about the refusal to change is exactly where Part 2 lands. Phase 3 isn't just a description of how institutions filter information — it's a description of why they stop adapting. When the evaluative question shifts from "is this true?" to "does this sound right to people like us?" you lose the feedback mechanism that would tell you to change course. The party isn't refusing to change out of stubbornness. It's that the knowledge infrastructure around the party has made it genuinely difficult to hear the signal through the noise.
As a life long Republican (I can still remember Dan’s Bake Sale in Colorado in 1989 when I was 10) I second your statement that Trump filled a vacuum created by the sclerotic leadership of the party. There is a story from back in 2016 when Trump met with Paul Ryan. At the meeting Ryan pushed tax policy and other “normie” Republican policy positions. Trump reportedly said “no wonder Ryan and Romney lost” when leaving the meeting. Trump understood that “vibes” trump policy (pun unintended) in the current political world. As a gen Xer I’m hyper focused on the upcoming scheduled cuts to social security (per current law when the trust fund is exhausted in 2033) and the other fiscal calamities that look poised to strike in the 1930s. Two decades of vibe based politics have us sleepwalking off a fiscal cliff.
Dan's Bake Sale — that's a deep cut. Colorado memory for a different generation of us.
The Ryan-Trump story captures the asymmetry perfectly. Ryan was operating in the Republican version of Phase 3: tax policy as institutional common sense, the thing that "sounds smart" in donor meetings and think tank panels. Trump looked at the same electorate and saw that the party's knowledge infrastructure had stopped tracking what its own voters actually cared about. He didn't have better policy. He had a functioning feedback loop, even if it was mostly instinct and crowd-reading.
Your point about vibe-based politics and the fiscal cliff is where this gets scary. Phase 3 on the left filters out inconvenient empirical challenges to coalition orthodoxy. Phase 3 on the right has its own version: a politics built on affect and grievance with no mechanism for processing boring, actuarially certain problems like the Social Security trust fund. The 2033 exhaustion date isn't a vibe. It's math. And right now neither party's epistemic infrastructure is equipped for that conversation. On one side it threatens coalition commitments; on the other it's not entertaining enough to survive the attention economy.
That's what wild about the pattern this piece describes. It's not just that individual publications get defunded or analysts get pushed out. It's that the capacity for boring, consequential, empirically grounded policy thinking is eroding on both sides simultaneously, for different reasons but with the same result.
Your response is certainly reasonable which I greatly appreciate. But for me the crux of the matter is that left and center left parties, especially in the larger countries face a similar problem: loss of working class support to the populist right, and a growing dependence on well educated white collar workers and professionals. Did they all turn to cultural politics? Or is this a structural issue? But keep in mind that left and center left parties have always faced competition for working class votes. As far as climate change you may well be correct that the choices made were suboptimal, but I would also consider the way that Democrats in energy producing states are cross pressured and certain options are not therefore not politically viable. Finally, I would add that the claims of Trump voters to be motivated by value differences is undermined by their support for the most morally compromised politician since slavery was abolished.
These are three distinct points and I want to take them seriously rather than lump them together.
On the cross-national pattern — this is the strongest version of the structural argument and I think you're largely right. The decline of left parties' working-class support tracks across democracies with very different cultural politics. (e.g., Piketty's work on this is compelling.)
Where I'd push back slightly is on the either/or framing. The structural shift (deindustrialization, credentialization, the knowledge economy sorting) created the conditions. But the pace and depth of the rupture vary significantly across countries, and the cultural choices parties made accelerated or slowed the break. The Democrats didn't have to lose working-class Hispanics at the rate they did — that was partly a function of specific cultural positioning decisions that other center-left parties handled differently. Structure sets the table. Culture decides how fast you eat.
On energy-state Democrats — fair point, and it illustrates exactly the kind of cross-pressure that Phase 3 can't process well. A framework that treats climate skepticism as heresy has no room for the Democrat in western Pennsylvania who supports the party on healthcare and labor but watches her community get written off as a sacrifice zone. That's not a messaging problem. It's a coalition design problem. TLP was trying to name that, and it's not an easy task to my mind.
On the Trump moral compromise argument, I think this one cuts less than you want it to. You're conflating voter motivation with candidate endorsement. Saying that working-class voters had substantive reasons to leave the Democratic Party doesn't require defending their choice of vehicle. People can be right about the diagnosis and wrong about the cure. The relevant question for Democrats isn't "how could they vote for him" — it's "what did we do that made him look like the better option?" That's the question Phase 3 makes it hard to ask, because it frames the answer as inherently illegitimate.
Well, its reasonable to question whether Democrats support for cultural positions is as significant as TLP claims. After all Democrats have been losing support from working class Americans since the 1980's. A trend we also see throughout the Western world among left and center-left parties. Perhaps the decline of unions and a political economy favorable to union growth is more significant. Also attributing a 1.5 percent defeat in 2024 to any particular cause ignores again incumbents lost everywhere in 2024. IMO, TLP rode their own hobby horses to explain a phenomenom better explained by inflation, immigration, and Democratic leaders unwillingness to tell Biden it was time to go. When over 60% of Democrats indicated they did not want him to run for re-election. Only the last is particular to Americans. It may be cultural issues show a distance between the professional class and workers, but whether it carries the weight being claimed is doubtful. Also the claim that climate change has been catastrophized is not one laymen can reasonably adjudicate.
You're raising the right competing explanation, and I think Teixeira would actually agree with more of this than you'd expect. Some of the working-class realignment predates the cultural turn — the union decline, deindustrialization, the Democrats' neoliberal drift in the '90s all opened the door. His argument (especially in Where Have All the Democrats Gone? with Judis) isn't that culture replaced economics as the explanation. It's that the cultural turn slammed shut a door that was already closing. The question is why Democrats couldn't get those voters back once they started talking about economics again, and his answer is that the cultural baggage made the economic pitch non-credible.
On 2024 specifically — the margin was narrow, and yes, incumbents lost everywhere. But the demographic shifts underneath it aren't cyclical. Hispanic support didn't halve because of inflation. Young men didn't swing 29 points because Biden was old. Those are structural moves that the incumbency frame can't explain on its own.
On the climate point — I'd push back on "not one laymen can reasonably adjudicate." Whether the 1.5-degree target is scientifically arbitrary versus well-supported is a question with a literature, and the Breakthrough Institute work TLP published was engaging that literature directly. "Trust the experts" is often the right call, but "you're not qualified to evaluate this" is also how Phase 3 polices its boundaries, I'd argue?
While TLP did go 10% beyond where I think the facts support on their "voters don't care about climate change"/"climate is the problem" thesis (e.g. https://guidedcivicrevival.substack.com/p/dont-over-interpret-the-red-wave), threatening their funding over their stance on climate was absolutely anti-intellectual and indicative of the pathologies of recent big-D Democratic coalitional politics, for all the reasons you and others have pointed out. (And TLP was 100% right about 1.5 being a bad target, all of the above being popular and the right approach, etc. Which is why I'm only saying they went 10% too far.). But regardless of how correct their interpretation of the facts was, pulling TLP support is super counterproductive to the donor's cause. Much better to have the argument on the facts. Shutting down debate always makes it look like--to everyone else watching--that you think you would lose the debate.
P.S. To be clear: I also think TLP was right in its general analysis of where the Democrats went wrong on culture, etc., and it was one of my favorite Substacks. I just had that one tiny quibble with their no-climate-voters thesis.
The 10% is fair, I think! (Teixeira sometimes collapsed "voters don't prioritize climate" with "climate policy is wrong," and those are different claims with different evidentiary standards.)
Your piece on not over-interpreting the red wave made that case well. But the 10% is what makes the defunding so revealing. The donor didn't engage the 10% where TLP was arguably overextended. They shut down the whole operation over the 90% that was landing. "Here's where your climate read goes too far" is a coalition doing its job. "Technically correct but harmful" is a coalition that's replaced empirical evaluation with utility evaluation. And you're exactly right that everyone watching notices the difference.
(And re: your PS — noted and agreed. One of mine too.)
You're right of course...the 1930s usage is a great explainer.
"Politically correct" originally meant "aligned with the party line"...truth evaluated by its utility to the movement, not by its correspondence to reality. That's literally the epistemological shift I'm describing. (We just don't call it that anymore because the phrase got strip-mined by the culture war, the Red Scare, etc.)
Great piece! And, oh man, the whole paragraph around this: "They’re people whose social identities, professional networks, and moral self-understanding are organized around being the kind of person who cares about climate, who stands on the right side of history, who supports the arc of progress." And, I think one could do some really cool political science on that insight that nicely expands existing social identity frameworks.
The identity-protection mechanism is doing a lot of work in the piece and I think it's genuinely undertheorized in the political science literature. We have good frameworks for partisan identity (Mason, Huddy), for elite cue-taking (Zaller), for motivated reasoning (Kunda, Kahan). But the specific dynamic where professional epistemic identity becomes the thing being protected, where what's at stake isn't "my party" but "my status as a serious, informed person," that's a different animal. It operates upstream of partisanship. You can see it in how academics sort which findings are "interesting" versus "problematic" before any political content enters the picture.
Someone should be writing the paper that maps how knowledge-class professional identity interacts with information evaluation in ways that aren't reducible to partisanship or ideology. There's probably a survey experiment in there: give people identical empirical claims sourced from high-status versus low-status institutions and measure not just belief updating but willingness to engage. I'd bet the identity threat shows up as avoidance rather than disagreement. People would say "that's not serious."
If you write it I want an acknowledgment footnote.
There's plenty of people who criticize the Democratic Party and want it to moderate who are quite successful - Matt Yglesias for example. Hell, Andrew Sullivan is a Tory who continually complains about Democrats but is doing quite well. To point at a new-ish publication, The Argument has many of the same criticisms of the Democratic Party, but seems to be doing much better.
The difference is they actually talk about when Democrats, even leftie Democrats do good things and yes, also point out when the Republicans do terrible things. You need some honey to get the medicine to go down.
The Liberal Patriot had one story - the Democrats are too woke and left-wing and they're going to lose because of this, and it turned out, even among people upset with the Democratic Party's current trajectory, there was a limited audience for a 100% anti-Democratic Party tone and you could see that in the comment sections, which were only trolls or people obsessed with anti-woke viewpoints, which is just as outside of the median Democrat as some left-wing bluehair in NYC.
I'm a dirty leftie so I was never going to agree w/ them, but there were multiple people who were moderates who pointed out the issues w/ the TLP model. Ruy as a voice in a broader center-left/centrist publication might've worked, but in shocking news, if your only story is "Democrats suck and need to do better," when even in Texas, people like Kamala have +74 approval ratings among Democrats, there's not much of a constituency outside of the very small amount of obsessive anti-woke centrists, who are disproportionately on sites like this and Twitter but are frankly out of touch with the median Evelyn Normielib as lefties may be out of touch w/ the median voter.
https://x.com/admcrlsn/status/2041354424891666653
This is a fair critique of TLP as a publication and I don't disagree with most of it. The tone did become pretty one-note, the comment sections were rough, and you're right that @Matthew Yglesias and @The Argument do similar work with more range and more success. If I were writing a media analysis of why TLP failed as a business, your explanation would be most of the story.
But that's not quite the argument I'm making. The piece is about what happened before TLP existed — when Teixeira was a senior fellow at CAP, Brookings, and AEI, publishing data-driven analyses of coalition realignment. The donor defunding and institutional exit preceded the Substack, and the Substack's tone probably reflects someone who'd already been pushed out of the spaces where nuance gets rewarded. That's not an excuse for the one-note quality — it's a sequencing point. The publication you're describing was partly a product of the pattern I'm describing.
On the constituency question — I'd actually agree that pure "Democrats are too woke" content has a limited audience, and that audience skews weird in exactly the way you describe. But Teixeira's early work wasn't that. The Emerging Democratic Majority was the most optimistic Democratic coalition analysis of its era. The shift from that to what TLP became is itself part of the story.
The comparison to Yglesias is interesting because it kind of proves both our points. Matt threads the needle — he'll criticize Democrats on process and strategy while clearly being on the team. He's also independently wealthy and doesn't depend on institutional funding. That independence matters more than tone, I think.
I think the contrast between Yglesias, the Argument, and TLP is really interesting. I love MY and pay him money and so far am holding out on the Agument because I already have too many subscriptions, but also love it. And, I often read TLP and appreciated it, but never to the same degree. And I think it's because TLP needed to expand beyond "Democrats suck" in ways that Yglesias and Demsas consistently do. Also, it's not like MY is operating off a trust fund, this needle he threads is lucrative. There's a market for smart, centrist criticism that gets that Dems on trans issues are dumb, but Trump on almost everything is 10x worse.
I think the market comparison is actually revealing in a way that supports both the above point and mine. Yglesias and Demsas succeed because they do the full range. They'll tell you Democrats are wrong on housing and then spend the next post explaining why the Republican alternative is worse. TLP didn't do that, and the publication suffered for it. No argument there.
But I think the interesting question is why TLP didn't do that. Some of it is just editorial choice and temperament. But some of it, I think, is what happens to your analytical range after you've spent years being told your empirical observations are unwelcome. Yglesias built his platform from a position of institutional security and social capital. Teixeira built his after being pushed out of the institutions where that range would have been natural. The Substack was the product of the pattern, not just a parallel case.
The market point is real though. There clearly is a big audience for smart center-left criticism that maintains team membership. The fact that Yglesias can be as heterodox as he is while staying legible as "one of us" is partly skill and partly the financial independence to not care if a donor flinches. That second part is underrated in these comparisons.
So here was the mystery to me.. of all the pieces Teixera wrote about various D aligned issues- why was climate the straw that broke the proverbial camel’s back. The effective policy “Sine qua non”? Could it be that “climate” and its only D-preferred solutions somehow are fundamental in a way that we are unable to discern? If we go back to the Hartwell paper in 2009, there have been a variety of solutions and priorities to decarbonizing, but only a narrow range are acceptable to the donor group. We could ask them “why?” except donors are not accountable to citizens.
The Hartwell Paper reference is sharp, and it's worth unpacking for people who don't know it. That paper (2010, out of LSE and Oxford, authored by 14 climate and social scientists including Pielke, Hulme, and Rayner) argued that the Kyoto framework had "dangerously narrowed the option space" for realistic climate policy. Their core claim was that treating climate as a discrete problem to be "solved" through one pathway (emissions caps, rapid fossil fuel phaseout, net zero targets) was both scientifically reductive and politically suicidal. They proposed multiple parallel approaches — low carbon taxes funding R&D, tackling black carbon and tropospheric ozone, forest protection, universal energy access — on the theory that decarbonization would only succeed as a contingent benefit of goals that were actually politically attractive, not as a moral crusade demanding sacrifice. (Tagging @rogerpielkejr, maybe he'll have something to say about this...).
So, yes, that's Phase 1 thinking applied to climate. Empirically grounded, pragmatic, treating the problem as persistent and complex rather than as a test of moral seriousness. And the Democratic donor class's response to it over the last 15 years has been almost exactly what you'd predict: the narrow pathway became orthodoxy, the Hartwell alternatives got filed under "not serious," and the range of permissible climate positions contracted until even TLP's relatively modest deviations were enough to trigger defunding.
Your question about why climate specifically was the tripwire is the one the piece probably underexplores. My read is that climate is where identity and policy are most tightly fused for this donor class. Immigration is a position you hold. Climate is a vocation. It organizes their philanthropy, their social networks, their sense of moral purpose. The other issues could be debated because they were held at arm's length. Climate couldn't be debated because it was held close. Questioning it didn't feel like a policy disagreement. It felt like an accusation.
And you're right that donors aren't accountable to citizens. That's part of what makes the Shadow Party so effective as a Phase 3 enforcement mechanism — it operates outside democratic feedback loops entirely.
First, let me state how much I'm enjoying your newsletter, which I found through someone linking to your "Reality Bats Last" post; you should have more than a few hundred subscribers!
As a Democrat for 50 years now, I'm sympathetic to your diagnosis, and find it interesting to contemplate how it applies to other issues, most notably education and crime. In the area of education I'm interested in reforms in states like Alabama and Louisiana that are apparently successful in helping low-SES students, as well as new tech-enabled education models like Alpha School, Math Academy, etc. However I suspect the chances of either having an effect on public education in my own state of Maryland are slim to none, given that they threaten both the prevalent orthodoxy on how public education should work and the economic interests of a major part of the Democratic coalition, namely public school teachers and related education professionals.
On the other hand, it's interesting to see Democratic officials in heavily-Democratic cities like San Francisco pursue policies to crack down on crime and public disorder, an issue that's traditionally GOP-coded. Perhaps that's because the costs of crime and disorder for that part of the Democratic coalition consisting of city-dwelling professionals are perceived to be high enough to outweigh the influence of other parts of the coalition that have advocated for restorative justice, prison abolition, defunding the police, and similar policies.
Thank you — that's very kind, and I'm glad "Reality Bats Last" brought you here!
Your two examples are doing exactly what I hoped the piece would prompt — applying the Phase 3 lens to other policy domains and seeing where it clarifies things.
The education case is a great illustration of Phase 3 operating at full strength. You've got empirical results from Alabama and Louisiana that would ordinarily interest anyone who cares about low-SES student outcomes — which is supposedly the whole point. But the results come wrapped in policy frameworks that threaten both the orthodoxy (how learning should be organized) and the coalition (who gets to organize it). So the evaluative question quietly shifts from "is this working?" to "does this empower the right institutions?" That's the utility-evaluation move I describe in Part 2. The evidence doesn't get refuted. It gets filed under the wrong heading and ignored.
The San Francisco crime case is interesting because it might actually be the exception that clarifies the rule. When the costs of a policy position land directly on the coalition members who hold institutional power — city-dwelling professionals dealing with visible disorder in their own neighborhoods — the feedback loop still works. The signal gets through because the people receiving it are the same people who set the agenda. That's not Phase 3 breaking down. That's Phase 3 never fully forming on that issue, because the constituency bearing the costs had enough institutional leverage to keep the empirical question live.
Which raises an uncomfortable implication: the coalition is most responsive to evidence when the professional class is the one paying the price. When the costs land on working-class members (lost energy jobs, rising insurance premiums from climate policy, school systems that don't work) the Phase 3 filter holds. That's not a conspiracy. It's just how social sorting interacts with institutional power.
Being a Canadian makes me somewhat hesitant to comment, but your question "what did we do that made him look like the better option?" made me stop and think.
First, it can be equally applied to the Republican Party as well. What did they do to allow Trump to take over the party?
From the outside, it seemed that Trump told the voters he understood they were unhappy and he would bring real change.
I think the lack of change in the Democratic Party is the answer to the original question. Hence, why their refusal to change is probably the most dangerous thing they can do.
All perspectives welcome here at the BBQ. :)
And you're right that the question applies to the GOP too. The Republican establishment's failure to take working-class economic grievances seriously created the vacuum Trump filled. The difference is that Republicans adapted (badly, in ways that traded principle for power) while Democrats are still debating whether the grievances are real. That's the asymmetry that worries me.
Your instinct about the refusal to change is exactly where Part 2 lands. Phase 3 isn't just a description of how institutions filter information — it's a description of why they stop adapting. When the evaluative question shifts from "is this true?" to "does this sound right to people like us?" you lose the feedback mechanism that would tell you to change course. The party isn't refusing to change out of stubbornness. It's that the knowledge infrastructure around the party has made it genuinely difficult to hear the signal through the noise.
As a life long Republican (I can still remember Dan’s Bake Sale in Colorado in 1989 when I was 10) I second your statement that Trump filled a vacuum created by the sclerotic leadership of the party. There is a story from back in 2016 when Trump met with Paul Ryan. At the meeting Ryan pushed tax policy and other “normie” Republican policy positions. Trump reportedly said “no wonder Ryan and Romney lost” when leaving the meeting. Trump understood that “vibes” trump policy (pun unintended) in the current political world. As a gen Xer I’m hyper focused on the upcoming scheduled cuts to social security (per current law when the trust fund is exhausted in 2033) and the other fiscal calamities that look poised to strike in the 1930s. Two decades of vibe based politics have us sleepwalking off a fiscal cliff.
Dan's Bake Sale — that's a deep cut. Colorado memory for a different generation of us.
The Ryan-Trump story captures the asymmetry perfectly. Ryan was operating in the Republican version of Phase 3: tax policy as institutional common sense, the thing that "sounds smart" in donor meetings and think tank panels. Trump looked at the same electorate and saw that the party's knowledge infrastructure had stopped tracking what its own voters actually cared about. He didn't have better policy. He had a functioning feedback loop, even if it was mostly instinct and crowd-reading.
Your point about vibe-based politics and the fiscal cliff is where this gets scary. Phase 3 on the left filters out inconvenient empirical challenges to coalition orthodoxy. Phase 3 on the right has its own version: a politics built on affect and grievance with no mechanism for processing boring, actuarially certain problems like the Social Security trust fund. The 2033 exhaustion date isn't a vibe. It's math. And right now neither party's epistemic infrastructure is equipped for that conversation. On one side it threatens coalition commitments; on the other it's not entertaining enough to survive the attention economy.
That's what wild about the pattern this piece describes. It's not just that individual publications get defunded or analysts get pushed out. It's that the capacity for boring, consequential, empirically grounded policy thinking is eroding on both sides simultaneously, for different reasons but with the same result.
Your response is certainly reasonable which I greatly appreciate. But for me the crux of the matter is that left and center left parties, especially in the larger countries face a similar problem: loss of working class support to the populist right, and a growing dependence on well educated white collar workers and professionals. Did they all turn to cultural politics? Or is this a structural issue? But keep in mind that left and center left parties have always faced competition for working class votes. As far as climate change you may well be correct that the choices made were suboptimal, but I would also consider the way that Democrats in energy producing states are cross pressured and certain options are not therefore not politically viable. Finally, I would add that the claims of Trump voters to be motivated by value differences is undermined by their support for the most morally compromised politician since slavery was abolished.
These are three distinct points and I want to take them seriously rather than lump them together.
On the cross-national pattern — this is the strongest version of the structural argument and I think you're largely right. The decline of left parties' working-class support tracks across democracies with very different cultural politics. (e.g., Piketty's work on this is compelling.)
Where I'd push back slightly is on the either/or framing. The structural shift (deindustrialization, credentialization, the knowledge economy sorting) created the conditions. But the pace and depth of the rupture vary significantly across countries, and the cultural choices parties made accelerated or slowed the break. The Democrats didn't have to lose working-class Hispanics at the rate they did — that was partly a function of specific cultural positioning decisions that other center-left parties handled differently. Structure sets the table. Culture decides how fast you eat.
On energy-state Democrats — fair point, and it illustrates exactly the kind of cross-pressure that Phase 3 can't process well. A framework that treats climate skepticism as heresy has no room for the Democrat in western Pennsylvania who supports the party on healthcare and labor but watches her community get written off as a sacrifice zone. That's not a messaging problem. It's a coalition design problem. TLP was trying to name that, and it's not an easy task to my mind.
On the Trump moral compromise argument, I think this one cuts less than you want it to. You're conflating voter motivation with candidate endorsement. Saying that working-class voters had substantive reasons to leave the Democratic Party doesn't require defending their choice of vehicle. People can be right about the diagnosis and wrong about the cure. The relevant question for Democrats isn't "how could they vote for him" — it's "what did we do that made him look like the better option?" That's the question Phase 3 makes it hard to ask, because it frames the answer as inherently illegitimate.
Well, its reasonable to question whether Democrats support for cultural positions is as significant as TLP claims. After all Democrats have been losing support from working class Americans since the 1980's. A trend we also see throughout the Western world among left and center-left parties. Perhaps the decline of unions and a political economy favorable to union growth is more significant. Also attributing a 1.5 percent defeat in 2024 to any particular cause ignores again incumbents lost everywhere in 2024. IMO, TLP rode their own hobby horses to explain a phenomenom better explained by inflation, immigration, and Democratic leaders unwillingness to tell Biden it was time to go. When over 60% of Democrats indicated they did not want him to run for re-election. Only the last is particular to Americans. It may be cultural issues show a distance between the professional class and workers, but whether it carries the weight being claimed is doubtful. Also the claim that climate change has been catastrophized is not one laymen can reasonably adjudicate.
You're raising the right competing explanation, and I think Teixeira would actually agree with more of this than you'd expect. Some of the working-class realignment predates the cultural turn — the union decline, deindustrialization, the Democrats' neoliberal drift in the '90s all opened the door. His argument (especially in Where Have All the Democrats Gone? with Judis) isn't that culture replaced economics as the explanation. It's that the cultural turn slammed shut a door that was already closing. The question is why Democrats couldn't get those voters back once they started talking about economics again, and his answer is that the cultural baggage made the economic pitch non-credible.
On 2024 specifically — the margin was narrow, and yes, incumbents lost everywhere. But the demographic shifts underneath it aren't cyclical. Hispanic support didn't halve because of inflation. Young men didn't swing 29 points because Biden was old. Those are structural moves that the incumbency frame can't explain on its own.
On the climate point — I'd push back on "not one laymen can reasonably adjudicate." Whether the 1.5-degree target is scientifically arbitrary versus well-supported is a question with a literature, and the Breakthrough Institute work TLP published was engaging that literature directly. "Trust the experts" is often the right call, but "you're not qualified to evaluate this" is also how Phase 3 polices its boundaries, I'd argue?
While TLP did go 10% beyond where I think the facts support on their "voters don't care about climate change"/"climate is the problem" thesis (e.g. https://guidedcivicrevival.substack.com/p/dont-over-interpret-the-red-wave), threatening their funding over their stance on climate was absolutely anti-intellectual and indicative of the pathologies of recent big-D Democratic coalitional politics, for all the reasons you and others have pointed out. (And TLP was 100% right about 1.5 being a bad target, all of the above being popular and the right approach, etc. Which is why I'm only saying they went 10% too far.). But regardless of how correct their interpretation of the facts was, pulling TLP support is super counterproductive to the donor's cause. Much better to have the argument on the facts. Shutting down debate always makes it look like--to everyone else watching--that you think you would lose the debate.
P.S. To be clear: I also think TLP was right in its general analysis of where the Democrats went wrong on culture, etc., and it was one of my favorite Substacks. I just had that one tiny quibble with their no-climate-voters thesis.
The 10% is fair, I think! (Teixeira sometimes collapsed "voters don't prioritize climate" with "climate policy is wrong," and those are different claims with different evidentiary standards.)
Your piece on not over-interpreting the red wave made that case well. But the 10% is what makes the defunding so revealing. The donor didn't engage the 10% where TLP was arguably overextended. They shut down the whole operation over the 90% that was landing. "Here's where your climate read goes too far" is a coalition doing its job. "Technically correct but harmful" is a coalition that's replaced empirical evaluation with utility evaluation. And you're exactly right that everyone watching notices the difference.
(And re: your PS — noted and agreed. One of mine too.)
Political correctness (in the 1930s usage, not the US right-wing media usage) is going to kill us all.
You're right of course...the 1930s usage is a great explainer.
"Politically correct" originally meant "aligned with the party line"...truth evaluated by its utility to the movement, not by its correspondence to reality. That's literally the epistemological shift I'm describing. (We just don't call it that anymore because the phrase got strip-mined by the culture war, the Red Scare, etc.)