PART TWO: No Offramp

The Failed Experiment and the Bad Equilibrium

At the end of Part One, I left you with a picture of an information environment that has been systematically degraded over decades — first by legitimate grievances against genuine governmental malfeasance, then by a right-wing media apparatus that turned distrust into a product, and finally by a president who made the unfalsifiable conspiracy theory the native language of American political discourse.

The obvious question is whether there's a way back.

The honest answer, arrived at by looking at what's actually happened rather than what we'd prefer to believe, is: not anytime soon. And the evidence is the Biden administration.

The Natural Experiment We'd Rather Forget

Biden ran on institutional restoration. Not just implicitly — it was the explicit thesis of his campaign and his presidency. The argument was: elect me, and I will demonstrate that normal, competent, norm-respecting governance is both possible and worth having.

The purest expression of that project was Merrick Garland at the Department of Justice. Garland was selected specifically for his reputation as a careful institutionalist. His restraint was meant to be visible — indeed to demonstrate, by example, that the DOJ could function as a neutral instrument of law rather than an extension of presidential will.[1]

What actually happened is that Garland's principled restraint was experientially indistinguishable from failure. The January 6th investigation moved so slowly that by the time charges were filed against Trump, they landed inside the 2024 campaign window — which handed the Republican apparatus exactly the framing it needed. Not "a president is being held accountable for trying to overturn an election," but "the weaponized DOJ is interfering in an election." The delay didn't prevent the politicization. It enabled it.

Here's the genuinely depressing part: Garland wasn't wrong, exactly. He was applying the right norms (note: norms that the Never Trump cohort expounds upon endlessly) to an environment where the norms had already been destroyed. You cannot demonstrate the value of institutional restraint to an audience that has been conditioned for two decades to see institutional restraint as either weakness or complicity.

The electorate's revealed preferences in 2024 weren't subtle. Economic grievance was the loudest signal, but running beneath it was something worse for the institutionalist thesis: process doesn't satisfy. Outcomes do. Institutional integrity that produces no accountability for norm violation is, from the perspective of someone who watched four years go by without anyone answering for anything, indistinguishable from institutional failure.

Biden proved that you can govern with fidelity to norms and lose anyway. That's a hell of a result for anyone who believed trust could be rebuilt top-down through demonstrated good behavior.

The Local Trust Caveat Doesn't Save You

A common observation — and not a wrong one — is that while trust in federal institutions has cratered, trust in local institutions remains substantially higher. Fire departments, local schools, local government. The argument goes that this surviving substrate of local trust could, in theory, provide the foundation for rebuilding upward.

It's hopeful. It's also doing too much work.

The data is an average. It averages over communities where local institutional trust is real and reasonably high, and communities where it has never been real — where the local institution in question is a police department with a decades-long documented history of violence and zero accountability. For Black communities in Louisville and Minneapolis and Memphis, for Latino communities in Phoenix, for poor communities whose encounters with local law enforcement have been defined by brutality rather than service, the "local trust is higher" claim is either false or irrelevant to their actual lived experience.

Their distrust of local institutions is not conspiratorial. It is the appropriate response to documented evidence, and their lived or observed reality.

And the current administration is actively making it worse.

The Trump DOJ's rollback of consent decrees — canceling the agreements with Minneapolis and Louisville reached after the killings of George Floyd and Breonna Taylor, while simultaneously closing investigations into departments in Phoenix, Memphis, and several other cities — is not just a policy reversal. It is trust destruction in exactly the communities that had the most legitimate reasons for distrust.

The theory behind consent decrees was modest: if a police department has a documented pattern of constitutional violations, put it under external monitoring and measure compliance. It was never fast, never clean. But it was the primary mechanism by which systematic misconduct got documented and subjected to scrutiny rather than buried.

What you get when the monitoring is removed is reversion to the norm.[2] The bad behavior wasn't corrected under the consent decrees in many cases — it was submerged. Bide your time, perform compliance, and return to the prior behavior once the monitors are gone. The communities that experience that cycle were already skeptical of the federal government's sincerity. Harmeet Dhillon's framing — that the decrees represented "anti-police" federal overreach — sends those communities a message that their documented experiences of abuse are not real, not the government's problem, or subordinate to the preferences of the departments that committed the abuse.

That's not trust-building. That's trust-destroying, conducted in the places where trust was already most fragile.

What It Would Actually Take

If upward trust flow from local to federal were going to happen, the conditions are specific and none of them are currently present.

First: demonstrated federal competence on something tangible and visible. The early COVID vaccine rollout in 2021 was briefly this — logistics working, doses moving, a visible federal effort producing a visible result. Trust ticked up. Then messaging chaos around boosters and masks erased most of that credit. Also, the Delta variant that spread like wildfire in early summer 2021 didn't help. Trust takes sustained positive experience to build and a single failure to damage.

Second: personnel continuity and institutional insulation from political cycles. The parts of the federal government that retain the most credibility tend to be the ones where career civil servants have genuine independence. The current administration's aggressive use of Schedule F reclassification — removing civil service protections from tens of thousands of career employees and pushing out the people who know why the processes exist — is a direct attack on exactly this mechanism. You cannot build institutional trust in an institution redesigned to serve the current president's preferences rather than consistent statutory mandates.

Third, and most dauntingly: a political incentive structure that rewards trust-building rather than punishing it. This just does not exist. The media ecosystem that profits from distrust has no economic reason to stop. The politicians who benefit from conspiratorial epistemology have no electoral incentive to stop. And the electorate, as 2024 demonstrated, does not reliably reward institutional patience.

The Bad Equilibrium

The sunk cost framing that started this conversation is emotionally accurate, but the more precise concept is a bad equilibrium — a coordination failure. Everyone would be better off in a high-trust environment, but no individual actor has sufficient incentive to move first. The cost of unilateral norm-following in a low-trust environment is immediate and personal. The benefit of restored trust is diffuse and long-term. Politicians who respect institutions get outmaneuvered by those who don't. Media that reports carefully loses audience to media that performs outrage.

The table isn't just crooked. The incentive to keep it crooked is built into the game, and one side is focused on running the table.

What might change it? The most honest answer is a sustained series of crises severe enough that conspiratorial epistemology produces consequences its believers can feel directly — not consequences felt by communities that were already skeptical, those communities have been paying the price for years and it hasn't moved the needle. Consequences felt by the people who currently benefit from the dysfunction.

That's not a recovery plan. That's an observation about what historically precedes them.

The ante is in the pot. The hand has been played. The next one looks a hell of a lot like the last one, and the dealer isn't changing.

1 - if you spend time at The Bulwark, you will hear Sarah Longwell talk about how important it is for the DoJ in particular to return to being neutral.

2 - see what I did there?