Trump’s Loyalty Market Is Already Changing Institutional Behavior

The most dangerous market in Washington right now does not trade stocks, bonds or crypto. It trades fear.

I do not mean that as a metaphor for ordinary partisan hardball. Presidents always reward friends, punish enemies around the edges and use legal powers in ways their opponents call corrupt. The pardon power is broad. Regulators have discretion. Prosecutors can charge powerful people. Universities and media companies that take federal money or need federal approval are not exempt from law.

But a line gets crossed when the lesson sent to institutions is no longer “follow the rules” but “guess what the president wants before the rules are applied.” I think Donald Trump’s second term has crossed that line. Not completely. Not irreversibly. Courts still matter, and some investigations of Trump critics may have real legal predicates. Yet the pattern is now clear enough to shape behavior: allies can expect mercy, critics can expect exposure and institutions can buy down risk through visible accommodation.

Start with the mercy side of the ledger. On January 20, 2025, Trump granted clemency to roughly 1,500 people charged or convicted in connection with January 6 and commuted 14 sentences, including for prominent extremist-group figures, according to NPR’s account of the proclamation¹. That was not a quirky end-of-term pardon for a donor or relative. It was wholesale relief for a class of offenders tied to a pro-Trump political event, including defendants convicted in cases arising from the attack on the Capitol.

The best defense of Trump here is structurally true and morally weak: presidents have abused clemency before. Joe Biden pardoned Hunter Biden on December 1, 2024, after saying he would not, and the pardon covered federal offenses committed or potentially committed from January 1, 2014, through December 1, 2024, according to the Associated Press². Biden also commuted roughly 1,500 sentences and pardoned 39 people in December 2024, which AP described as the largest single-day clemency act in modern history, though that action covered people released to home confinement during the pandemic and people convicted of nonviolent crimes, according to AP’s report carried by PBS³. So yes, the pardon power is an old problem. But Trump’s January 6 clemency is different in kind because it converted loyalty-adjacent lawbreaking into a mass political amnesty.

That is the first price signal: if your illegal conduct can be framed as service to the leader’s cause, accountability may be temporary.

The pressure side is messier, but more important. James Comey, Letitia James and John Bolton are not interchangeable cases. Bolton, Trump’s former national security adviser turned critic, was indicted in October 2025 on classified-information charges, and several legal analysts told FactCheck.org⁸ the indictment looked more detailed and stronger than the cases against Comey and James. That matters. A Trump critic can commit a prosecutable offense. If every probe of an opponent is treated as persecution, the rule of law becomes impossible in the other direction.

Comey is the harder case for Trump’s defenders. He was indicted in September 2025 after Trump publicly pressed Attorney General Pam Bondi to move against political opponents, including Comey, Adam Schiff and Letitia James, and Reuters reported that career prosecutors had drafted a memo warning that the evidence did not establish probable cause, according to Reuters⁴ and ABC News⁵. Then a federal magistrate judge found a “disturbing pattern of profound investigative missteps” in securing the Comey indictment, including alleged legal misstatements to the grand jury and irregularities in the transcript, according to AP via PBS⁶.

In November 2025, a federal judge dismissed the Comey and Letitia James cases after concluding that Lindsey Halligan, the prosecutor who brought them, had been unlawfully appointed; the ruling did not decide whether the underlying allegations were true, but it did set aside actions taken under what the court called a defective appointment, according to AP via PBS⁷. That distinction cuts both ways. It means Trump’s critics cannot honestly claim a court already found the cases fabricated. It also means Trump’s defenders cannot honestly say the system worked cleanly. The cases entered the bloodstream through an appointment process a federal court later deemed unlawful.

This is where the “loyalty market” idea matters. Markets do not require certainty. They require expected value. If the expected cost of angering the president rises enough, rational actors change behavior before a judge ever reaches the merits.

The institutional examples are the strongest evidence because they show adaptation in real time. Trump issued executive orders targeting major law firms, and some firms fought. Perkins Coie won a permanent injunction in May 2025 when a federal judge ruled Trump’s order unconstitutional and said the government may not use federal power to punish lawyers for representation and viewpoints it dislikes, according to CBS News⁹. Jenner & Block also won a permanent block against a similar order, with AP describing the order as unconstitutional retaliation designed to punish legal work the White House disliked, according to AP¹⁰.

Those court wins are real. They are also not the whole story. Paul, Weiss got a targeted order rescinded after agreeing to review its hiring practices and provide $40 million in free legal services supporting administration priorities, according to CBS News¹¹. By April 2025, major firms had pledged about $940 million in pro bono work, Axios reported, with some firms choosing deals while others went to court, according to Axios¹². A court can later vindicate the resisters. It cannot erase the signal sent to everyone watching: litigation may work, but tribute is faster.

Columbia University tells the same story in campus form. On July 23, 2025, Columbia announced a federal resolution agreement under which it would pay $200 million over three years, while a vast majority of grants terminated or paused in March 2025 would be reinstated and access to current and future federal grants restored, according to Columbia’s own statement¹³ and the agreement page¹⁴. The Trump White House framed the settlement as a victory for enforcing anti-discrimination law and changing elite higher education, according to its fact sheet¹⁵. Columbia framed it as a way to end damaging uncertainty. Both can be true. The practical lesson for other universities is simpler: funding can disappear first, negotiated compliance can restore it later.

Paramount supplies the media version. In July 2025, Paramount agreed to pay $16 million to settle Trump’s lawsuit over CBS’s editing of a “60 Minutes” interview with Kamala Harris, according to Axios¹⁶. Later that month, the FCC approved the Paramount-Skydance merger; CNBC reported that FCC Chair Brendan Carr cited “significant changes” at CBS and Skydance commitments on viewpoint oversight and diversity, equity and inclusion policy, while the approval came less than a month after the Trump settlement, according to CNBC¹⁷. I cannot prove a quid pro quo from public evidence. I do not need to. The public sequence is enough to teach every regulated media owner that presidential grievance may become merger risk.

The strongest counterargument is that this is all too impressionistic. Prosecutors still need indictments. Courts have blocked Trump orders. Bolton may be a legitimate case. Columbia and Paramount may have settled for ordinary business reasons. Biden’s pardons show that ugly clemency did not begin in 2025.

I take that seriously. It keeps me from saying the justice system has been fully converted into personal rule. But it does not save the narrower and more important point. A loyalty market is not a world where law disappears. It is a world where law remains on the books while access, timing, relief and pain increasingly depend on political position. The danger is not only that Trump may punish enemies. It is that everyone else starts punishing themselves in advance.

The indicator to watch is not whether one spectacular Trump action loses in court. Several already have. The indicator is whether institutions keep choosing concessions over litigation when legal claims look contestable. If, by the end of 2026, more universities, law firms and media companies accept policy changes, payments or governance restraints to resolve politically charged federal threats, the market will have matured. If major institutions instead begin winning fast injunctions and refusing settlements without suffering lasting funding, licensing or client losses, I will revise my view. For now, the price board is visible, and too many powerful actors are learning to trade on it.