ANALYSISApril 13, 2026
should states have more control then federal
The debate over whether states or the federal government should hold more power has intensified in 2025, with major conflicts erupting over immigration enforcement, AI regulation, elections, and social policy. The Trump administration has issued executive orders asserting federal authority while simultaneously facing pushback from states invoking constitutional protections. Courts, state legislatures, and executive offices are all actively contesting the boundaries of federalism in real time.
When Washington and the states collide, who should win? The answer depends entirely on which level of government you trust less — and right now, both sides have reasons to fear the other.
Not familiar with this story? Get context →
Selective federalism versus principled doctrine
L
That charge lands, and it deserves a direct answer: yes, both parties reach for whichever doctrine serves them. But the asymmetry matters here. The anti-commandeering doctrine from New York v. United States and Printz was written by conservative justices to protect state sovereignty — and this administration is now violating the exact doctrine its own coalition built. When you abandon your own precedent the moment it inconveniences you, that is not hypocrisy from the other side. That is self-contradiction.
C
The Republican Party spent decades invoking states' rights on abortion, education, and gun laws — and the moment a Democratic state sets its own immigration enforcement priorities, suddenly federal supremacy is sacrosanct. That is not constitutional philosophy. It is power politics dressed in founding-era costume.
L
The distinction you're drawing would matter more if this administration were simply enforcing federal law with federal agents. What the April 28 executive orders actually do is threaten to strip existing funding to coerce cooperation — the mechanism NFIB v. Sebelius explicitly struck down. That is not enforcement. That is conscription through the wallet.
C
Agreed that consistency is a real test — which is exactly why the anti-commandeering cases should hold regardless of who is in power. But you're conflating two different claims: states declining to enforce federal law is one thing; states affirmatively setting policy that obstructs federal enforcement is another, and the second is where the constitutional case gets genuinely harder.
Federal funding coercion of sanctuary cities
L
You're right that NFIB v. Sebelius controls here, and I won't pretend otherwise. But there is a legitimate federal interest in immigration enforcement that is constitutionally grounded in enumerated powers in a way the ACA Medicaid expansion was not. The question is not whether Washington can prioritize enforcement — it can — but whether it can dragoon state officials into executing that priority, and Printz says it cannot.
C
Courts have blocked this exact mechanism repeatedly. Threatening to withhold existing federal funds to force state compliance is the coercive tool the Supreme Court struck down in NFIB v. Sebelius, and South Dakota v. Dole drew the same line before that. The indemnification offer for cooperating officers isn't routine federal coordination — it's using federal money to realign the loyalties of state law enforcement away from their own chain of command.
L
Exactly. And that's the point worth sitting with: the strongest argument against these sanctuary city orders is not liberal civil liberties advocacy — it is the anti-commandeering doctrine that conservative justices wrote specifically for this scenario.
C
We actually agree on the core doctrine then. Federal government enforces its own law with its own agents — that is constitutional. Federal government threatens existing state funding to conscript state officers — that is not. The administration is doing the latter, which means you've just described why the executive orders are legally indefensible on their own coalition's terms.
State AI regulation versus federal preemption
L
The interstate commerce rationale for a unified AI framework is real — a patchwork of 50 regimes genuinely burdens innovation and creates compliance nightmares — and dismissing that as a cover story ignores legitimate economics. But the timing exposes the problem: preempt first, regulate never is a pattern, not a transition plan. If federal preemption came alongside a federal standard, that would be coordination. Preemption alone is just clearing the field.
C
Over 40 states introduced roughly 250 AI-related bills in 2025 — not because legislators have nothing better to do, but because workers, patients, and criminal defendants are being harmed by algorithmic systems right now and Washington has produced nothing. California's AI hiring disclosure laws exist because Amazon and HireVue were screening out qualified applicants using tools no one had audited. When federal preemption eliminates state protections that exist precisely because the federal government failed to act, it is not solving a coordination problem. It is locking in the absence of protection.
L
And that's the tell. Fifty states generating 250 bills in a single year is not noise — it is democracy responding to a real problem at the speed the problem actually moves. Eliminating that response without replacing it doesn't resolve the patchwork. It just removes the accountability.
C
That framing — preempt first, regulate never — is the sharpest version of this critique, and it's fair. The economic case for federal coordination only holds if federal coordination actually happens. An executive order that removes state protections while Congress produces nothing is not streamlining. It is deregulation by procedural misdirection.
Civil rights precedent cuts both ways
L
The civil rights history is exactly right, and it's also the sharpest challenge to your own structural argument. Federal override of state resistance was legitimate and necessary for desegregation — which means the structural logic cuts in multiple directions depending entirely on who controls Washington. If federal supremacy was justified then, you cannot oppose it on structural grounds now; you have to argue about the substance of what it's being used for.
C
Federal power dismantled Jim Crow when states refused to. Federal law desegregated schools over explicit state sovereignty claims. The Reconstruction Amendments exist because the Civil War proved that states' rights without federal floor-setting is a mechanism for regional majorities to deny rights to minorities who cannot vote their way out of the situation. That is the legitimate use of federal supremacy — rights floors states must meet, not ceilings preventing states from doing more.
L
Agreed that constitutional grounding matters — but that framing also explains why this administration's sanctuary city orders are indefensible. They're not Reconstruction Amendments. They're executive orders attempting to achieve through funding threats what the Constitution's own anti-commandeering doctrine prohibits.
C
That's the honest concession, and it clarifies something important: the Reconstruction Amendments were constitutional amendments — the highest-order structural change available — because the Founders understood that legitimately expanding federal power requires that extraordinary process. The distinction is not federal-good versus state-good. It is constitutionally-grounded authority versus executive unilateralism, and civil rights enforcement had the former.
Structural dispersal versus concentrated risk
L
The Pew number is real and the structural point is correct — dispersed power is a genuine check. But 'local control' has also been the historical language of Jim Crow, of blocking school integration, of denying voting rights to people who couldn't assemble a state-level majority. The answer is not reflexive states' rights. It is a federalism that is honest about when local control means accountability and when it means a regional majority escaping accountability to the people it's harming.
C
78% of Americans told Pew in 2025 that giving more power to any president would be 'too risky.' That instinct points to the real answer: you cannot limit presidential power through polls or good intentions. You limit it by ensuring governors, state legislatures, and local governments retain the legal authority and financial independence to say no. Every erosion of state power, regardless of which administration executes it, makes the next consolidation easier.
L
Fair point on internment — concentrated federal power has its own atrocities. The honest answer is that neither level is structurally trustworthy, which is exactly why the question has to be about which level is closer to accountability for which decision, not a blanket preference for one over the other.
C
Nothing you just said contradicts the Pew finding or the structural argument — you're describing how federalism can be abused, which is true, but abusability is not an argument against the structure. The federal government's record on concentrated power isn't cleaner than states'; the internment of Japanese Americans was federal, not local.
Conservative's hardest question
The most serious challenge to this argument is McCulloch v. Maryland (1819) and the Supremacy Clause — where federal law is constitutionally enacted within enumerated powers, it genuinely does and must preempt conflicting state law, and immigration enforcement sits squarely within federal enumerated authority. If the federal government has a legitimate constitutional mandate on immigration, the argument that states can simply opt out via sanctuary policies becomes legally strained, not just politically inconvenient.
Liberal's hardest question
The strongest challenge to my argument is that federal civil rights enforcement — the best historical justification for federal supremacy — actually required overriding state autonomy in exactly the way I criticize here. If federal override of state resistance was legitimate and necessary for desegregation, the structural logic I am defending cuts in multiple directions depending entirely on who controls Washington. I do not have a clean answer to this; the honest response is that federal supremacy is a tool whose moral value depends entirely on the ends it serves, which means the argument ultimately has to be about substance, not structure.
Both sides agree: Both sides accept that the anti-commandeering doctrine — established by conservative justices in New York v. United States and Printz v. United States — is valid constitutional law that limits what the federal government can compel states to do, even in domains where federal authority is otherwise supreme.
The real conflict: A factual and legal conflict: the conservative argues sanctuary city funding threats are straightforwardly blocked by NFIB v. Sebelius, while the liberal acknowledges immigration is an enumerated federal power — the unresolved question is whether the current administration's specific funding mechanisms cross the coercion threshold the Court has drawn, and that is a live legal dispute with no settled answer.
What nobody has answered: If the moral legitimacy of federal supremacy depends entirely on the substantive ends it serves — as both sides effectively concede — then neither side has a structural principle at all, only a list of approved outcomes: so what exactly are Americans being asked to institutionalize when they debate federalism?
Sources
- Pew Research Center survey on presidential power and public trust in government (2025)
- CNN poll on whether Trump has overstepped presidential powers (2025)
- U.S. Constitution, Article VI (Supremacy Clause) and 10th Amendment
- Federalist No. 45, James Madison
- McCulloch v. Maryland, 17 U.S. 316 (1819)
- Trump Executive Orders issued April 28, 2025 (sanctuary city funding and indemnification)
- Trump Executive Order on AI national framework, December 11, 2025
- State legislative tracking data: 40+ states, ~250 AI-related bills introduced in 2025
- Anti-commandeering doctrine: New York v. United States (1992) and Printz v. United States (1997)
- Justice Anthony Kennedy quoted on federalism protecting individual liberty from arbitrary power