Module 2: The U.S. Dual Court System: State & Federal

Every U.S. state has its own complete court system, and there is also a separate, parallel federal court system — both operating within the same geographic territory, deciding different (sometimes overlapping) categories of cases. Neither system sits above the other as a general matter; they interact only at defined points.

Heath v. Alabama, 474 U.S. 82 (1985), shows just how genuinely separate these two systems really are — separate enough that the same conduct can be prosecuted twice, once by each. Heath was convicted of murder in Alabama state court after arranging his wife’s kidnapping and killing, which occurred partly in Alabama and partly across the border in Georgia; Georgia then separately prosecuted him for the same underlying murder. Heath argued the second prosecution violated the Double Jeopardy Clause’s bar on being tried twice for the same offense. The Supreme Court rejected that argument, applying the dual sovereignty doctrine: because Alabama and Georgia are each independent sovereigns, each with its own distinct interest in enforcing its own criminal law, a crime committed within both states’ territory constitutes two separate offenses for double jeopardy purposes, not one. The same logic extends to state and federal prosecutions of the same conduct. Heath is a genuinely vivid illustration of this section’s core point — state and federal systems aren’t one unified hierarchy that treats a single act as one single legal event, they are truly separate systems each capable of independently vindicating its own distinct legal interest in the same underlying conduct.

Example. A single dispute arising in Ohio might be heard entirely in Ohio’s own state courts, entirely in federal court, or — depending on how it’s filed and whether it qualifies — potentially shift from one system to the other.

The federal system has three basic tiers: U.S. District Courts (trial level), U.S. Courts of Appeals organized into geographic circuits (appellate level), and the U.S. Supreme Court (highest level, largely discretionary review) — a direct, concrete illustration of Module 1’s general framework.

Not every one of the thirteen federal circuits is purely geographic. Twelve circuits — the eleven numbered regional circuits plus the D.C. Circuit — hear appeals from district courts within their own defined territory. The thirteenth, the Federal Circuit, is genuinely different: created by the Federal Courts Improvement Act of 1982, it has no geographic territory of its own at all, instead hearing appeals nationwide in specific subject-matter categories — most notably patent cases, along with certain claims against the federal government and international trade disputes. Congress created this subject-matter-based exception specifically to address a real, documented problem: patent law had been developing inconsistently across the different regional circuits, with the same patent potentially treated as valid in one circuit and invalid in another, creating genuine unpredictability for inventors and businesses. Concentrating patent appeals in one specialized, nationwide court was a deliberate structural fix for that specific inconsistency problem.

Example. Several states typically share the same federal circuit — a party appealing a federal district court decision appeals to that circuit’s Court of Appeals, not to a court dedicated to a single state alone.

Most states roughly mirror this tiered structure, but real naming variation exists — even within a single country. Most famously, New York’s trial-level court of general jurisdiction is actually called the “Supreme Court,” while New York’s actual highest court is called the Court of Appeals — the reverse of what someone unfamiliar with New York’s system might reasonably assume.

This isn’t an arbitrary quirk — it traces to New York’s own Constitution of 1846, which created the “Supreme Court” as the name for the state’s general trial court of unlimited original jurisdiction, following a naming convention common in many state constitutions of that era where “supreme” described the breadth of a court’s subject-matter authority (able to hear virtually any type of case) rather than its position at the top of an appellate hierarchy. New York’s actual highest appellate court, the Court of Appeals, was established separately in that same 1846 constitution specifically as the state’s final appellate authority. New York is far from unique in having unusual court names, but it remains the most commonly cited example specifically because its trial-level “Supreme Court” sits directly opposite what nearly every other U.S. state calls its own highest court, creating exactly the kind of confusion this section warns comparative and interstate practitioners to watch for.

Example. A case described as being before the “New York Supreme Court” has not necessarily reached New York’s highest court at all — it may still be at the trial level, with a full appeal and a further appeal to the Court of Appeals both still ahead of it.

Federal question jurisdiction gives federal courts jurisdiction over cases arising under the U.S. Constitution, federal statutes, or treaties — grounded in the nature of the legal claim itself, not the parties’ citizenship.

American Well Works Co. v. Layne & Bowler Co., 241 U.S. 257 (1916), gave this section’s “nature of the claim” test its classic, most-quoted formulation. American Well Works sued a competitor in state court, alleging the competitor had wrongly told customers American Well Works’s pump infringed the competitor’s patent — a state-law business-disparagement claim, not a patent infringement claim itself. The competitor argued the dispute really turned on patent validity, a federal question, and should be heard in federal court. Justice Oliver Wendell Holmes, writing for the Court, rejected that argument with a famously terse formulation: “A suit arises under the law that creates the cause of action.” Because the plaintiff’s actual claim was created by state disparagement law, not federal patent law, the suit arose under state law even though a federal patent question might come up somewhere along the way in defending it. Holmes’s formulation remains the starting point for federal question analysis today — the focus stays on what law actually created the plaintiff’s own claim, not on whether some federal issue might surface incidentally within the case.

Example. A claim alleging a violation of a specific federal consumer-protection statute invokes federal question jurisdiction regardless of whether the plaintiff and defendant happen to live in the same state.

Diversity jurisdiction lets federal courts hear certain state-law cases between citizens of different states, requiring both complete diversity (no plaintiff shares citizenship with any defendant) and an amount in controversy exceeding $75,000.

Diversity jurisdiction’s real justification goes back to the Constitution’s own ratification debate. In Federalist No. 80, Alexander Hamilton defended giving federal courts this power specifically by pointing to a genuine, practical worry: an out-of-state litigant — particularly an out-of-state creditor trying to collect a debt — might reasonably fear that a local state court, staffed by locally elected or appointed judges (and, especially in that era, local juries), could favor the local resident over the outsider. Hamilton argued that offering a neutral federal forum, insulated from that specific local-bias risk, was essential to give out-of-state parties genuine confidence in interstate commercial dealings, and by extension to the young nation’s broader economic integration across state lines. The Judiciary Act of 1789 then implemented this protection directly, creating diversity jurisdiction as one of the federal courts’ original bases of authority — a genuinely old, deliberately designed feature of the system, not a later add-on, even though its practical necessity today remains a real, live debate this module returns to.

Example. A California citizen suing a Texas citizen in a $200,000 purely state-law contract dispute likely satisfies both requirements, even though no federal law issue is present at all.

Concurrent jurisdiction means a matter can be heard in either state or federal court; exclusive jurisdiction means a matter must be heard in only one system — federal courts, for example, generally have exclusive jurisdiction over federal crimes, patents, and bankruptcy. A related doctrine, supplemental jurisdiction, lets a federal court that already has proper jurisdiction over one claim also hear a closely related state-law claim from the same underlying facts.

Gunn v. Minton, 568 U.S. 251 (2013), shows the exclusive/concurrent line genuinely mattering, in a case involving exactly the patent jurisdiction this section flags. An inventor sued his own former patent attorney for legal malpractice in Texas state court, alleging the attorney had botched a patent-infringement case by failing to raise a timely argument, causing the inventor to lose his patent rights. The attorney argued the malpractice claim actually required deciding a substantial patent-law question (whether the missed argument would have won), which he said triggered federal courts’ exclusive patent jurisdiction, ousting the Texas state court entirely. The Supreme Court disagreed, holding that even though the malpractice claim required analyzing a hypothetical patent question, the underlying claim itself was still a state-law legal malpractice claim, and resolving the embedded patent question wouldn’t actually affect the validity of any real, existing patent — the Court reasoned federal exclusivity exists to protect a genuine, uniform national interest in patent rights themselves, an interest simply not at stake in one client’s malpractice suit against his own lawyer. Gunn shows the exclusive/concurrent boundary is drawn around real substantive impact, not merely whether a federal-law question happens to come up somewhere in the analysis.

Example. A federal-statutory claim paired with a closely related state-law fraud claim arising from the same transaction can generally proceed together in federal court through supplemental jurisdiction, rather than requiring two separate lawsuits in two separate systems.

Removal lets a defendant move a case originally filed in state court to federal court, if the case could have originally been filed there. And the two systems aren’t fully sealed off from each other even after judgment: the U.S. Supreme Court can, in certain circumstances, review a state court’s highest-court decision that genuinely raises a federal question.

Example. A defendant sued in state court over a claim that independently qualifies for federal diversity jurisdiction may generally remove the case to federal court, so long as removal is timely and procedurally proper.

Caterpillar Inc. v. Lewis, 519 U.S. 61 (1996), addresses a genuinely odd procedural wrinkle in how removal’s timing actually works. A case was removed to federal court on diversity grounds, but at the moment of removal, complete diversity technically didn’t yet exist — one non-diverse party remained in the case. That remaining defect was later cured (the non-diverse party settled out before trial), and the case proceeded to a federal jury verdict without anyone raising the original defect again until after judgment. The Supreme Court held that because complete diversity existed by the time judgment was actually entered, and no party was prejudiced by the earlier procedural defect, the improper removal didn’t require throwing out an otherwise valid federal judgment reached after a full trial. Caterpillar is a genuinely practical, real-world reminder that removal’s jurisdictional requirements, while real and enforceable, are not always treated as absolute in every procedural posture — courts have shown some real willingness to avoid wasting a full trial’s worth of proceedings over a jurisdictional defect that no longer exists and harmed no one by the time it actually mattered.