Thomas B. Edsall writes in the New York Times:
In January 2025, when the U.S. House took up legislation to bar trans women’s participation on women’s sports teams, all but two Democratic representatives — Henry Cuellar and Vicente Gonzalez — voted against the bill.
When the Senate took up a similar proposal three days ago, every Democrat present voted against it.
Why don’t more Democrats explicitly moderate their stands on transFgender rights, immigration and other issues? Those who maintain far-out positions are well to the left of the electorate and its emblematic median voter. The trans issue clearly weakened Kamala Harris’s presidential campaign, leaving her open to devastating pro-Trump ads.
In the case of one of the most disputed rights claimed by some parts of the transgender activist community — transgender women’s participation on women’s sports teams — Democrats have clear liberal grounds to challenge that claim, by asserting that they are protecting a woman’s right from unfair competition.
But this phenomenon — drifting far from the median voter — is hardly limited to the left. There are many factors behind the reluctance of both Democrats and Republicans to shift to the center.
For one thing, donors, especially the growing legions of small donors, prefer more extreme candidates. Adding additional pressure, what have come to be known as “the groups” — advocacy organizations on the left and the right — demand fealty to policies that are sometimes politically costly; they threaten to support primary challengers to run against those who defy their authority. On a psychological level, Democrats and liberals are morally committed to protecting marginalized groups from harm and defending racial and sexual minorities…
Ruy Teixeira, a political analyst and senior fellow at the American Enterprise Institute, took a long look at this development in a March 12 posting on the Liberal Patriot Substack “The Democrats’ White Liberal Problem”:
Cast your mind back to the beginning of the century. At that point, a mere 28 percent of Democrats described themselves as liberal and two-thirds were either moderate or conservative.
Fast forward to today and the liberal share has more than doubled, to 59 percent, while the moderate/conservative share has declined drastically. It’s the liberals’ party now. And especially, it’s the white liberals’ party now.
How have white liberals changed?
In 2000, white Democrats who were moderate or conservative outnumbered white liberal Democrats by about two to one. Today that relationship has been reversed. White liberal Democrats now outnumber moderate/conservative white Democrats by about two to one.
The result: The balance of power within the party has moved in a decisively leftward direction:
From being merely a voice, albeit an important one, in the Democratic choir, white liberals are now directing the choir and imposing their culture, preferences and priorities on the party as a whole.
Any Democrat seeking the presidential nomination, Teixeira continued,
has to reckon with this enormous bloc of Democrats, whose influence is enhanced beyond their considerable numbers by their dominance of the party’s infrastructure, allied NGOs and advocacy groups, and left-leaning media, foundations and academia. Not to mention the money — ambitious Democrats need money, and white liberals are a reliable source of cash for politicians who press the right buttons.
This clarifies why it is so difficult for Democratic politicians to carve out a truly moderate path.
What else pushes Democrats to the left? Cash.
In their July 23, 2025, Wall Street Journal article, “AOC, Mamdani and Progressives Rake In Cash as Democrats Remain Divided: Far Left’s Prolific Fund-Raising Shows Appeal to Party’s Base,” John McCormick and Anthony DeBarros wrote:
Among the 10 incumbent Democrats who raised the most from individual donors this year, six are members of the Congressional Progressive Caucus, a Wall Street Journal analysis of campaign finance disclosures shows. Three of the top four are progressives, with the exception of House Minority Leader Hakeem Jeffries (D., N.Y.).
The financial strength among progressives presents a challenge to party leaders trying to nudge the Democratic message closer to the middle, where they might stand a better chance of winning over independent voters who decide close elections.
This is a clean application of my power framework to a concrete political problem, and the Edsall piece gives me good empirical scaffolding to climb.
Edsall identifies the structural math correctly. Safe districts remove the electoral incentive to moderate. Primary voters reward ideological intensity. The 17 percentage point advantage for extreme candidates in primaries is a striking figure, and the fact that 40 percent of state legislative races go uncontested is arguably the more damning number. Two-party competition, the mechanism supposed to force candidates toward the center, simply does not exist in a large fraction of American elections. Edsall is right about all of this. What he does not fully explain is why the actors inside these institutions experience themselves as reasonable while everyone outside experiences them as untethered from reality.
My framework supplies the missing piece. Edsall describes external incentives. What he leaves out is the internal epistemology, the fact that inside elite institutions, reasoning does not precede legitimacy. Legitimacy precedes reasoning. You signal alignment, you gain standing, and then you reason within the bounds that standing permits. The vocabulary, affirming care, harm reduction framework, centering equity, systemic inequities, is not descriptive. It is a credential. It tells other insiders that this person understands the rules and can be trusted with institutional authority. Once those signals are in place, the reasoning that follows can be quite sophisticated. People cite studies, build models, write careful memos. From the inside it feels like serious inquiry. But the constraint is already baked in. Certain conclusions are simply off the table because reaching them would violate the signaling structure that grants legitimacy in the first place.
This is why the gap with the median voter keeps widening without producing much visible distress among professionals. The median voter optimizes for outcomes: cost of living, crime rates, fairness in competition, border conditions. The professional class optimizes for coalition maintenance under moral constraints. These are different games. Edsall shows that safe districts reward ideological intensity at the electoral level. What he does not show is that institutions reward moral conformity at the professional level regardless of electoral outcomes. The double lock means that even a Democrat representing a genuinely competitive district faces institutional costs for moderating that have nothing to do with her primary electorate. Moderating might help her win in November. It might also cost her fellowships, donor relationships, media access, and speaking invitations. Those are not trivial losses.
Ruy Teixeira’s observation about white liberals is essential here and deserves more weight than Edsall gives it. White liberals now outnumber moderate and conservative Democrats by two to one, a complete reversal from 2000. This group controls the party’s infrastructure, its allied NGOs, its donor networks, its academic and media relationships. They do not merely vote. They staff, fund, and manage the institutions through which the party’s moral vocabulary gets produced and enforced. A candidate who violates that vocabulary does not just face a primary challenge. She faces a much broader and more diffuse form of exclusion from the class that runs the party’s machinery.
At the top of the hierarchy the language changes entirely, which is the most important thing Edsall misses. Senior donors and party brokers do not perform moral alignment because their status is already secured through pedigree and track record. They talk about risk exposure, institutional stability, and blowback. They manage the gatekeeping through hiring, funding, and the quiet allocation of access. By the time an argument reaches public debate, the selection has already happened. The conversation is downstream of the gatekeeping. This is why the reasoning can look polished but the conclusions feel predetermined to outsiders. The sorting happened upstream. The public argument is the retail operation. The wholesale operation already determined who gets to speak.
The transgender issue illustrates all three layers cleanly. At the retail layer, activists and mid-level professionals produce the moral vocabulary and enforce it through social pressure and institutional complaint mechanisms. At the mid-tier, Democratic politicians in safe districts calculate that violating the vocabulary costs more in institutional standing than it gains in general election votes, which for most of them never comes into question anyway. At the wholesale layer, major donors and party strategists are watching the polling on trans issues in competitive districts with considerable alarm but communicating their concerns privately rather than publicly, because public dissent would violate the norms of coalition solidarity that maintain their own standing. The result is a system where nearly every Democrat in the House votes against a bill that clear majorities of the public support, and where almost none of them experience this as a problem because the optimization problem they are solving is not the one the median voter thinks they are solving.
Mara Keisling’s call for retrenchment is interesting precisely because it comes from someone with established credentials inside the coalition. She can say what she says without losing standing because her pedigree is secure. A junior staffer or a first-term representative from a competitive district who said the same thing would face a very different reception. That asymmetry is the system working exactly as my framework predicts.
