The Erie Doctrine Explained: Which Law a Federal Court Applies (1L Civ Pro)

The Erie doctrine answers a single question: when a federal court hears a state-law claim (almost always because it has diversity jurisdiction), does it apply state law or federal law? The short answer is that it applies state substantive law and federal procedural law.

The doctrine comes from Erie Railroad Co. v. Tompkins (1938), which overruled Swift v. Tyson and held that there is no "federal general common law." A federal court in diversity must apply the substantive law of the state in which it sits, including that state's judge-made common law.

The reason Erie is hard is that the easy part — apply state substantive law — immediately runs into a hard part: telling substance from procedure. Most of the doctrine is a series of tests for drawing that line when it isn't obvious.

Before Erie, under Swift v. Tyson, federal courts in diversity applied their own "general" common law instead of state common law. That produced two problems the Erie Court cared about: forum shopping and the inequitable administration of the laws.

Forum shopping meant that the outcome of a case could depend on whether you sued in state or federal court — so a party with access to federal court (often an out-of-state defendant) could get a different, sometimes better, rule of law. That undermines the idea that like cases should be treated alike.

Erie tied these concerns to the structure of federal power: there is no federal authority to make general common law for matters the Constitution leaves to the states. So in diversity, the federal court's job is to predict and apply state law, not to invent its own.

After Erie, courts needed a way to classify a given rule as substantive (apply state law) or procedural (apply federal law). Three cases built the framework, each refining the last.

York's outcome-determinative test was powerful but too broad — almost any difference can affect the outcome. Byrd softened it by adding a balancing step: even an outcome-affecting state rule can yield when a strong federal interest is at stake. Hanna then reorganized the whole analysis into two clean tracks.

Hanna splits the analysis based on whether a Federal Rule of Civil Procedure (or a federal statute) directly conflicts with the state rule. This is the decision tree most 1L exams want you to run.

Track one — a Federal Rule is on point: if a valid Federal Rule of Civil Procedure answers the question, the federal court applies it, even if state law differs. Under the Rules Enabling Act, a Federal Rule is valid as long as it is arguably procedural and doesn't "abridge, enlarge, or modify" a substantive right. In practice, the Federal Rules almost always survive this step.

Track two — no Federal Rule on point: if there's no Federal Rule directly on point, you apply the "twin aims of Erie" — would applying federal law instead of state law (1) encourage forum shopping, and (2) lead to the inequitable administration of the laws? If yes, apply state law. The classic example is a state statute of limitations, which is outcome-determinative and invites forum shopping, so it controls.

A reliable exam move: first ask whether a Federal Rule (or statute) is directly on point. If yes, run the Rules Enabling Act / Hanna track. If no, run the twin-aims analysis. Naming which track you're in is half the credit.

The most common Erie error is invoking it in a case that isn't governed by state substantive law at all. Erie is about diversity cases (and the state-law claims in a case) — situations where the rule of decision comes from state law. In a pure federal-question case, federal substantive law already applies, and Erie's substance question doesn't arise the same way.

A second frequent mistake is jumping straight to the outcome-determinative test and forgetting Hanna. After Hanna, you don't start with York; you start by asking whether a Federal Rule is on point. The outcome-determinative idea lives inside the twin-aims track, not as the first question.

Finally, students sometimes treat "procedure" as a magic word that always means federal law wins. It doesn't — a rule can look procedural and still be outcome-determinative enough (like a limitations period) that Erie's aims require state law. Run the framework rather than labeling by intuition.