President Trump has now pardoned Tina Peters. Many commentators have rushed to say that this changes nothing. The president can pardon only federal crimes, they tell us. Peters was convicted under Colorado law; therefore, the pardon is merely symbolic. Perhaps. But suppose we treat the pardon not as theater but as the opening move of a serious constitutional argument. Suppose we ask whether the standard assumption is actually right. And suppose we ask a further question, one that matters far beyond Colorado. If a president can pardon only federal crimes, then states can neutralize the president’s constitutional power whenever they dislike its use. If the pardon power was designed to function as a national safety valve, that would be an odd design.
The steelman case is simple to state. Having issued the pardon, Trump should take Colorado into court and force the Supreme Court to answer a question that has been left in an awkward half-shadow for generations. When the Constitution grants the president power to grant reprieves and pardons for offenses against the United States, does it mean only violations of federal criminal statutes, or does it mean any criminal offense that, in substance, constitutes an attack on the national interest, even if a state happens to be the prosecuting sovereign? The Tina Peters case is an unusually clean vehicle for this dispute, because the whole fight is about the national electoral process, and because the conflict between the president and a state executive is stark and public.
Begin with the text. Article II grants the President the power to grant reprieves and pardons for offenses against the United States, except in cases of impeachment. The clause has one explicit limit, impeachment. That structure matters. When the Constitution wishes to carve out categories, it often does so directly. Here, it announces a broad power and then names a single exception. The common modern move is to treat offenses against the United States as a synonym for federal crimes. But that is an interpretive conclusion, not a piece of grammar. The phrase is capable of a more substantive reading. An offense against the United States is an offense against the United States, an act that wrongs the nation, undermines its constitutional order, or threatens the national polity. Federal criminal statutes are one way of identifying such acts, but not obviously the only way.
A puzzled reader may ask, how could a state offense be an offense against the United States? The answer is that the wrongness of an act is not determined by the label that a prosecutor puts on it. Think of an assault on a federal officer carrying out federal duties. A state can call it assault. The federal government can call it interference with federal functions. The same act, described at different levels, can implicate different sovereign interests. The pardon clause may be keyed to the interest, not the code section.
This is not an academic point about semantics. It goes to the function of the power. The framers did not place the pardon power in Article II as an ornamental flourish. They placed it there because they expected moments when law would be used in ways that threatened stability, reconciliation, or national unity, and when a single elected official would need authority to cut through the machinery. That is why the power is vested in one person, not parceled out among committees. A power designed to preserve the republic in moments of political heat is a power that must be able to operate in the very arenas where political heat is highest. Those arenas are not limited to federal indictments.
Early Supreme Court language points in the direction of breadth. In Ex parte Garland, the Court described the pardon power as unlimited except for impeachment and said it extends to every offense known to the law, and may be exercised at any time after its commission, before proceedings, during proceedings, or after conviction. The Court emphasized that the power is not subject to legislative control. The passage is not a careful treatise on federalism, and critics will say it presupposes federal offenses. But the point of citing Garland is not that it already resolves the present question. The point is that the Court’s own rhetoric has never treated the pardon power as a small, technical privilege confined to a narrow subset of prosecutable wrongs. The Court has repeatedly treated it as a large constitutional instrument.
United States v. Klein is even more revealing. In Klein, Congress attempted to manipulate the legal consequences of presidential pardons in the context of Civil War property claims. The Court rejected Congress’s effort in part because it impaired the effect of the pardon, and because the Constitution entrusted the pardon power to the executive alone, granted without limit. The Court used a famous phrase: a pardon blots out the offense pardoned and removes all its penal consequences. Again, the case did not involve a state prosecution. But it did involve the deeper idea that a pardon is meant to be decisive. A pardon is not an advisory memo. It is a constitutional act that changes legal reality.
Now bring in a case that shows the Court’s willingness to read offenses against the United States in a way that is not tethered to the federal criminal code. In Ex parte Grossman, the Court held that the president may pardon criminal contempt of court. Contempt is not, in the ordinary way of speaking, a federal statutory crime. It is an offense against the authority of the federal judiciary. Yet the Court treated it as within the pardon clause. The reasoning is important. It suggests that what matters is not whether Congress has placed a label in the U.S. Code, but whether the conduct counts as an offense against the sovereign authority of the United States.
At this stage, a cautious reader will raise the obvious objection. The clause says offenses against the United States, and the U.S. is a federal government of enumerated powers. States are sovereign in their own sphere. So surely state crimes are outside the pardon power. This objection is not frivolous. It is, in fact, the prevailing view, and it is defended in modern summaries of constitutional law. The steelman response, however, is that the objection assumes what it needs to prove. It assumes that the only way an offense can be against the United States is by being prosecuted by the United States. But the Constitution does not say prosecuted by. It says against.
And the Constitution’s own structure complicates the neat boundary. The Supremacy Clause makes valid federal law the supreme law of the land, binding judges in every state, notwithstanding anything in state law to the contrary. The president’s pardon power is valid federal law in the highest sense, a direct constitutional grant. If the president can constitutionally erase the legal consequences of an offense that is against the United States, then a state cannot nullify that erasure by recasting the same conduct as a state offense, at least when the national interest is the very interest that the state prosecution is targeting.
Notice the narrowness of the claim. The steelman argument does not require that presidents be able to pardon any state offense. It requires only that there exist a category of cases where state prosecution is, in substance, a prosecution of an offense against the United States, because the conduct implicates a core federal interest, and because the practical effect of the prosecution is to punish, deter, or stigmatize behavior in a domain where the national government has constitutional primacy. Election administration is a good example. States run elections, but federal elections select federal officers. The Constitution itself builds a federal structure around that fact, through Article I’s Elections Clause, through Article II’s presidential selection mechanism, and through Congress’s authority to set certain election rules. That makes the national interest in election integrity and in fair treatment of those who contest elections unusually direct.
This brings us to Tina Peters. Whatever one thinks of her conduct, the story that motivates Trump’s pardon is a national story. Peters was a county clerk. Her actions were taken, she and her supporters claim, in service of election integrity, and in response to widespread concern about the administration of the 2020 presidential election. The prosecution, and especially the severity of the sentence, has been widely interpreted on the right as political retaliation, not neutral law enforcement. One can steelman that interpretation without insisting on every factual claim in the dispute. The basic point is that Peters’ case sits at the intersection of state criminal enforcement and the national political process.
If that is right, then the constitutional question is not merely whether the president can reach into state court and tear up a conviction. The deeper question is whether state prosecutions can be used to accomplish what the Constitution denies other actors the power to accomplish, namely, to defeat the legal effect of the President’s clemency judgments in matters tied to the national interest. If states can always punish a pardoned person for the same underlying conduct by choosing state charges, then the federal pardon power is partly a paper tiger. It exists on paper but can be evaded in practice.
A skeptical reader may respond, dual sovereignty allows both federal and state prosecutions for the same conduct, so why could not a state punish even if the federal government has forgiven. But dual sovereignty is not a license for states to trench upon federal constitutional powers. It is a doctrine about prosecutorial authority in the absence of conflict. When conflict exists, the question becomes one of supremacy. If the president issues a pardon grounded in the national interest, and the state insists on continuing punishment precisely to negate that judgment, then the state is not merely exercising its own sovereignty. It is attacking an enumerated federal power.
There is also a practical problem with the narrow view that the Tina Peters episode brings into focus. Under that view, a president can pardon a federal defendant, and that pardon is legally powerful. But the president cannot protect similarly situated people whom a state chooses to charge for closely related conduct, even when the conduct is integrally bound up with a national political contest. This creates a perverse incentive. Partisan state prosecutors can weaponize state criminal law to punish the same set of politically disfavored actions that the president wishes to treat with mercy or restraint. If the president’s pardon power is meant to serve as a stabilizing national mechanism, it should not be so easily circumvented.
That is why Trump’s next move matters more than the pardon itself. A symbolic pardon is politics. A lawsuit is constitutional architecture. If Trump truly believes, as he has said, that Peters is being punished for political reasons and that her treatment threatens the integrity of national elections, then he should convert the dispute into a clear test of federal supremacy and constitutional meaning. The posture would be straightforward. Peters remains incarcerated under Colorado judgment. Trump has issued a pardon and declares that the continued imprisonment is unlawful because it conflicts with the federal clemency power as properly understood. Colorado responds that the pardon is void as to state convictions. The courts must then decide.
Some will insist the Supreme Court will reject the claim quickly. Perhaps it will. The dominant modern view is that the pardon power reaches only federal offenses. Even the most sympathetic justices will be wary of announcing an expansion that could be portrayed as federal takeover of state criminal justice. But steelmanning requires us to notice that the Court also has tools to craft a narrow, principled rule. It could hold that state convictions are ordinarily outside the pardon power, but that there is an exception where the state prosecution directly targets conduct bound up with federal functions, federal officers, or the integrity of federal elections, and where the president has made a formal determination that the prosecution undermines a national interest. That would preserve state autonomy in the vast run of ordinary cases while honoring the constitutional design that grants a unique national safety valve.
Another puzzled reader may ask, would this not allow presidents to declare any state crime a national interest and then pardon it? The steelman reply is that constitutional powers are often subject to political and institutional constraints even when they are formally broad. The president already has immense discretion in the pardon power as it is currently understood. A president can pardon friends, allies, and donors for federal crimes. That is a political check problem, not a textual limitation problem. And courts are capable of policing obvious bad faith when a doctrine is formulated with criteria that focus on federal functions and conflicts.
There is also an important symmetry here. Critics of Trump often argue that American democracy depends on letting states prosecute former officials, candidates, or political actors when federal prosecution is unavailable or politically inconvenient. The Tina Peters pardon fight is the mirror image. The right is arguing that American democracy also depends on protecting political dissenters from being crushed by state machinery when that machinery becomes a tool of faction. A constitutional system built to survive political conflict should be able to recognize both dangers.
Why insist on a Supreme Court ruling rather than leaving it to Colorado’s governor to pardon or commute? Because the point is not merely to free one defendant. The point is to clarify a constitutional boundary that, if left vague, invites gamesmanship. If the law is that state prosecutions can always defeat federal pardons, then presidents will either stop using the power in politically sensitive cases or they will seek other mechanisms of pressure that are cruder and more destabilizing. If the law is that federal clemency can, in narrow circumstances, preempt state punishment where a national interest is directly implicated, then the system gains a stabilizing rule rather than a perpetual standoff.
The Tina Peters case is a good vehicle precisely because it forces us to confront what the Constitution means by offenses against the United States. Federalism is not a slogan. It is a structure of overlapping authorities that sometimes collide. The pardon clause was written to give the national executive a tool for those collisions. It would be strange to read the clause in a way that makes it least effective when the stakes are highest.
Trump has already taken the first step by issuing the pardon. The second step is harder, but also more serious. Bring the case. Let Colorado defend the narrow reading. Let the Solicitor General defend the broader reading. Let amici from across the ideological spectrum test the arguments. And let the Supreme Court tell the country, in a reasoned opinion, whether the president’s power of mercy is a purely federal housekeeping device, or whether it is, as the founding design suggests, a national instrument meant to cut through jurisdiction when the constitutional order itself is at stake.
If you enjoy my work, please subscribe: https://x.com/amuse.
Sponsored by the John Milton Freedom Foundation, a nonprofit dedicated to helping independent journalists overcome formidable challenges in today’s media landscape and bring crucial stories to you.
READ NEXT: Good Laws Are Making These Conservative Victories Harder To Measure
boycott Colorado