A Critical Guide to Erie Railroad Co. V. Tompkins

Nelson, Caleb, William and Mary Law Review

II. JUSTICE BRANDEIS'S ARGUMENTS IN ERIE

Federal courts applied the doctrine described in the previous Part until 1938. In that year, however, Justice Brandeis's opinion in Erie instructed federal judges to follow state-court precedents on all questions that lie within the states' legislative competence, even if those questions would previously have been classified as matters of "general" law. (78) Brandeis left various details to be worked out in later cases. (79) But on the basic issue raised in Erie itself, Brandeis's opinion was clear: within the limits of the states' lawmaking powers, precedents that a particular state's highest court had established and to which it continued to adhere were to have the same status in federal court as statutes enacted by the state's legislature.

Justice Brandeis advanced three different categories of arguments for this conclusion. First, he argued that Justice Story's opinion in Swift v. Tyson had misinterpreted section 34 of the Judiciary Act of 1789 and that the original meaning of the statute compelled the result in Erie. (80) Second, he observed that "the doctrine of Swift v. Tyson" was having bad effects in practice. (81) Third, he argued that the course pursued by the federal courts under Swift was "unconstitutional[ ]" (and, seemingly relatedly, that it rested on a "fallacy" about the nature of law). (82) This Part examines each of those arguments in turn.

A. Justice Brandeis's Historical Argument

Justice Brandeis opened with an argument about history. As mentioned above, section 34 of the Judiciary Act of 1789 had specified that "the laws of the several states, except where the constitution, treaties or statutes of the United States shall otherwise require or provide, shall be regarded as rules of decision in trials at common law in the courts of the United States in cases where they apply." (83) (By the time of Erie, this statutory provision was known as the "Rules of Decision Act" (84) and was found at section 721 of the Revised Statutes of 1874. But section 721 of the Revised Statutes was substantively identical to section 34 of the original Judiciary Act, and Justice Brandeis's opinion therefore referred throughout to section 34.) In Swift v. Tyson, Justice Story had held that the phrase "laws of the several states" in section 34 covered (1) written state laws, (2) "long established local customs having the force of laws," and perhaps (3) settled precedents of the state courts about the content of these "local" laws, but not (4) decisions of the state courts on "questions of a more general nature," such as "questions of general commercial law." (85) According to Justice Brandeis, however, "the more recent research of a competent scholar" had "established" that "th[is] construction ... was erroneous" and that section 34 had been intended to make federal courts follow state-court precedents about the "general" aspects of the state's unwritten law as well as the "local" aspects, (86) The "competent scholar" whom Brandeis cited was Charles Warren, who had unearthed new information about the drafting history of section 34.s7 Warren had located ("in the attic of the Capitol") the original version of the Judiciary Act as introduced in the Senate, and he had also located ("in a cellar room, under a heap of miscellaneous papers of confused and intermingled dates and subjects") a copy of the bill as approved by the Senate and sent to the House. (88) Warren saw that section 34 did not appear in the former document but did appear in the latter document, meaning that it was added at some point during the Senate's consideration of the bill. (89) What is more, Warren found evidence of the amendment that added it. Among other "odd, loose slips of paper of different sizes and shapes" that set forth proposed amendments to the judiciary bill and that had been "preserved in a bundle, in the Senate Files [in the Capitol's attic]," (90) Warren discovered a sheet--apparently in the handwriting of Senator Oliver Ellsworth (91)--setting forth both a draft of what became section 34 and the edits that produced the final version. Although some words in the draft are struck out, they are still legible, and they show that the draft referred to "the Statute law of the several States in force for the time being and their unwritten or common law now in use, whether by adoption from the common law of England, the ancient statutes of the same or otherwise." (92) Through emendations that appear on the same sheet, this passage was replaced with the shorter phrase "the laws of the several States," yielding the provision that appears as section 34 of the Judiciary Act. From this progression, Warren inferred that "the [phrase] 'laws of the several States' was intended to be a concise expression anda summary of the more detailed enumeration of the different forms of State law, set forth in the original draft," (93) and hence that section 34 encompassed not only each state's "Statute law" but also its "unwritten or common law." For Warren, it followed that section 34 dictated the result that Justice Brandeis would later reach in Erie rather than the result that Justice Story had reached in Swift v. Tyson. Indeed, Warren wrote as if this conclusion were obvious: "Had Judge Story seen this original draft of the amendment, it is almost certain that his decision would have been the reverse of what it was." (94)

Even if one were otherwise to accept Warren's argument (and, as we shall see, that would be a serious mistake), the idea that information about the drafting history of section 34 would have changed the result in Swift v. Tyson is anachronistic. While the practice of consulting drafting history in statutory interpretation had become common by Warren's day, neither Justice Story nor his predecessors would have considered it appropriate. (95) Indeed, even modern judges who are sympathetic to this practice (96) might resist invoking the particular type of document unearthed by Warren--a sheet that (1) was not publicly available for the first 135 years after the Judiciary Act became law and (2) reflected a drafting change that might not have been widely known even among members of the enacting Congress. (97)

Still, these threshold objections are not themselves fatal to Warren's position. Even if the type of drafting history unearthed by Warren is not something that the early Supreme Court would have used to interpret early statutes, and even if it also is not the sort of thing that the modern Supreme Court would use in interpreting modern statutes, perhaps it is still useful when the modern Court needs to interpret an early statute--that is, when modern interpreters have to try to recreate the understandings of a past world. If Warren's arguments were valid, they might shed light on how at least one distinguished lawyer of the late eighteenth century used the phrase "laws of the several States." And unless we have some reason to think that Senator Ellsworth had a different understanding of that phrase than his contemporaries, perhaps this evidence sheds light on the original meaning of section 34 after all. (98) For the sake of argument, then, let us set aside any objections to Warren's willingness to consult drafting history in the first place, and let us also assume that whatever the drafting history shows about Senator Ellsworth's understanding of section 34 can properly be treated as the original meaning of that provision. Warren's argument still has two serious flaws.

The first flaw is well known to modern scholars. The crucial premise of Warren's argument is that when Senator Ellsworth or his colleagues decided to use the phrase "the laws of the several States" as a substitute for the earlier draft's reference to "the Statute law of the several States in force for the time being and their unwritten or common law now in use," the change was mostly stylistic rather than substantive: the shorter phrase was intended to refer to all the same types of law as the more detailed phrase. (99) As modern scholars have pointed out, though, Warren offered no support for this assumption, and the assumption is not obviously correct. (100) After all, legal draftsmen often change the language of a bill in order to alter its meaning, not to keep its meaning the same. Suppose, however, we could somehow know that Senator Ellsworth's changes were indeed primarily stylistic. There is still a second major flaw with Warren's argument--one that is less familiar to modern scholars than the first flaw, but more devastating. Suppose that section 34 had explicitly said what Warren took it to mean: "the Statute law of the several States ... and their unwritten or common law ... shall be regarded as rules of decision in trials at common law in the courts of the United States in cases where they apply" (except where the Federal Constitution, federal treaties, or federal statutes shall otherwise require or provide). Warren's conclusion still would not follow. In 1789, people did not automatically treat the phrase "unwritten or common law" as a synonym for "judicial decisions." (101) Thus, even if section 34 had explicitly instructed federal courts to draw rules of decision from the "unwritten or common law" in use in each state, section 34 would not necessarily have been ordering federal courts to accept state-court precedents about the content of the unwritten or common law. Instead, lawyers of the day could have read section 34 in the same way that Chancellor Taylor read the ordinance specifying that "the common law of England ... shall be the rule of decision" in Virginia: even if section 34 adopted the common law in use in each state, it did not necessarily adopt the decisions of each state's highest court. (102) More generally, section 34 need not be read to address the deference that federal courts owe to the decisions of each state's highest court about the content of any of the sources of law that section 34 invokes. (103)

Contrary to Justice Brandeis's claims, then, the drafting history uncovered by Charles Warren failed to show that Swift v. Tyson was "erroneous" as a historical matter, or that the original meaning of section 34 supported the result in Erie rather than the result in Swift. Modern scholars with an interest in history now generally agree that Warren's discovery was "inconclusive[]." (104)

The late Professor Wilfred J. Ritz would have taken this revisionism even farther: in his view, not only was Erie wrong about the original meaning of section 34, but even Swift may have yoked the federal courts too closely to the local law of individual states. Ritz began by observing that eighteenth-century legal draftsmen, including those responsible for the Judiciary Act of 1789, often used the phrase "the several states" to refer to "the states as a group" rather than to each state individually. (105) To convey the latter idea, Ritz argued, draftsmen preferred to use the phrase "the respective states" or "the states respectively." (106) Although Ritz acknowledged that this usage was not a "hard-and-fast rule," (107) he believed that section 34 reflected it: according to Ritz, when section 34 told federal courts to regard "the laws of the several states" as rules of decision, it was referring to "American law generally" rather than "the law of a particular state." (108) Ritz concluded that section 34 had not been intended to require federal courts to apply even the statutory law of any individual state (pace Justice Story's opinion in Swift), let alone what the courts of any individual state had said about the unwritten law of that state (pace Justice Brandeis's opinion in Erie).109 For other reasons, Ritz himself believed that section 34 was probably intended to address only criminal trials at common law and to have no bearing on civil cases. (110) But if section 34 did reach civil cases, Ritz took it to be an instruction to apply "the 'laws of the several states' viewed as a group of eleven states in 1789, and not viewed separately and individually." (111)

A number of distinguished modern scholars have accepted Ritz's view about the meaning of the phrase "the several states" in section 34.112 But the evidence does not really bear out that view. Even today, the adjective "several" can be used to refer serially to each discrete unit in a composite group. (Think of what it means to say that all the defendants in a civil case are being held liable not only "jointly" but also "severally.") This usage dates far back in American law, (113) and the First Congress often applied it to states--as when Congress resolved that the Secretary of State should "procure from time to time such of the statutes of the several states as may not be in his office," (114) or when Congress appropriated money "[i]or paying salaries to the late loan-officers of the several states," (115) or when Congress referred to "the requisitions heretofore made upon the several states." (116) The Constitution itself used the same locution when it provided that "[t]he House of Representatives shall be composed ofMembers chosen every second Year by the People of the several States." (117) Just as the phrase "the People of the several States" in this provision refers to the people of each state individually rather than the people of the states as an undifferentiated mass, the standard reading of section 34 takes the phrase "the laws of the several states" to refer to the law of each state individually. Not only is that reading consistent with the drafting habits of the late eighteenth century, but I am not aware of any persuasive evidence that Ritz's contrary reading of section 34 even occurred to a single lawyer or judge in the early Republic. By contrast, there is clear evidence of lawyers and judges adopting the standard reading. (118)

But while I think we can safely conclude that Ritz was wrong about the original meaning of section 34, that conclusion does nothing to rehabilitate Justice Brandeis's historical argument in Erie. Even though section 34 is naturally understood to refer to the laws of each state individually, and even though that reference can readily be understood to include unwritten as well as written forms of law, section 34 still need not be interpreted to address whether federal courts must defer to each state's highest court about the content of the unwritten law in force within that state. At any rate, the drafting history unearthed by Charles Warren has no real bearing on this point: it does not establish--or even suggest--that the original meaning of section 34 foreclosed Justice Story's position in Swift.

B. Justice Brandeis's Practical Arguments

Aside from making a specious historical argument, Justice Brandeis also offered practical reasons to dislike the doctrine associated with Swift. Here, he was on somewhat sounder ground, although he overstated his case and ignored plausible counterarguments.

1. The Murkiness of the Distinction Between "General" and "Local" Law

To begin with, Justice Brandeis observed that "the distinction between questions of general and of local law," which was central to the Swift regime, had proved both arbitrary and murky: no natural "line of demarcation" separated the two types of questions, and the federal courts' efforts at classification had introduced a "well of uncertainties." (119) That is true. But as students of civil procedure can confirm, Erie has not eliminated all line-drawing problems. According to Suzanna Sherry, indeed, Erie "simply traded one set of uncertainties for another," because it requires federal courts to draw equally murky distinctions between "substantive" questions (which Erie potentially instructs federal courts to answer in accord with the precedents of a particular state's supreme couro and "procedural" questions (which, in federal court, are instead governed by some form of federal law). (120)

Evaluating Professor Sherry's specific argument about the distinction between "substance" and "procedure" is harder than one might think, because federal courts had to draw that distinction even before Erie. One of the main reasons why they needed to do so, however, was a federal statute called the Conformity Act, (121) which was largely superseded in September 1938 and completely repealed in 1948. (122) Given the demise of the Conformity Act, modern federal courts may indeed have to distinguish "substance" from "procedure" somewhat more often because they follow Erie than would be necessary if they still followed Swift. (123) But to the extent that this conclusion depends on legal changes that occurred after Erie (such as the repeal of the Conformity Act in 1948), blaming Justice Brandeis for it may be unfair.

Even apart from the distinction between "substance" and "procedure," though, there is a different reason why Professor Sherry is correct that Erie "traded one set of uncertainties for another." (124) In fields over which the states enjoy legislative jurisdiction, Erie did eliminate the need for federal courts to separate questions of "general" law from questions of "local" law. But wherever federal courts following Swift would have characterized a question in one of those fields as "general" (and hence not controlled by the decisional law of any individual state), federal courts now have to identify the particular state whose decisional law matters. As a result, federal courts following Erie face horizontal conflict-of-law questions substantially more often than they did under Swift--and the lines that conflict-of-law analysis requires courts to draw may be just as murky as the line between '10cal" and "general" law)2"

At first glance, the Supreme Court's 1941 decision in Klaxon Co. v. Stentor Electric Manufacturing C0. (126) might seem to mitigate this problem. In one sense, after all, Klaxon established a clear rule: in cases covered by Klaxon, each federal district court is supposed to borrow the conflict-of-law doctrines that would be used by the highest court of the state in which it sits. (127) But unless that state has achieved clarity in conflict-of-law analysis, this approach does not eliminate the need for federal courts to draw murky lines; it simply tells federal courts to try to apply the same murky lines that the state courts would use. This instruction, moreover, has its own costs. Indeed, it arguably destroys one of the principal practical advantages that the diversity jurisdiction of the federal courts used to serve. (128)

2. Disuniformity and Forum Shopping

Aside from the line-drawing necessitated by Swift, Justice Brandeis also pointed to a more serious problem. As Part I suggested, people of Justice Story's day may well have hoped that the practice reflected in Swift would promote national uniformity on questions of general law; although the courts of individual states were not bound to accept what the Federal Supreme Court said about the content of the general law, people hoped that they would usually choose to do so (either because they found the Supreme Court's opinions persuasive or simply because those opinions provided a natural coordination point). (129) According to Justice Brandeis, however, this hope had not been realized, or had been realized only very imperfectly: state courts often "[p]ersist[ed] ... in their own opinions on questions of common law.' (130) In at least some states, then, the state courts applied different rules of decision than the federal courts on some questions of general law. The result was that "[i]n attempting to promote uniformity of law throughout the United States, the doctrine [associated with Swift] ... prevented uniformity in the administration of the law of the State." (131)

This disuniformity had various bad consequences. Justice Brandeis himself emphasized the incentives that it created for what is now called "forum shopping": lawyers contemplating litigation on behalf of a client sometimes knew that a state court would take a different view of the applicable law than a federal court located in the same state, and they sometimes could maneuver their client's case into the court whose view favored their side. (132) But forum shopping was an artifact of a more fundamental problem that had bad effects at an even earlier stage. In Henry Hart's words, the arrangements reflected in Swift "subject[ed] citizens at the crucial level of everyday activity to dual and often inconsistent systems of substantive law, without means of foretelling which system, in the unforeseeable contingency of litigation, was going to apply." (133) If state and federal courts have different understandings of the legal rights and duties associated with conduct in the real world, and if someone who might engage in that conduct cannot know in advance which court system will adjudicate any lawsuits arising from his conduct, the resulting uncertainties may well produce both inefficiencies and injustices. (134)

Still, as various scholars have explained, when we consider whether Erie is better than Swift in this respect, we need to keep in mind two different sorts of uniformity (and two different sorts of disuniformity). (135) Swift offered the prospect of uniformity on questions of general law across all federal courts throughout the country: once the United States Supreme Court had addressed and resolved a question of general law, all other federal courts were …

Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: A Critical Guide to Erie Railroad Co. V. Tompkins. Contributors: Nelson, Caleb - Author. Journal title: William and Mary Law Review. Volume: 54. Issue: 3 Publication date: February 2013. Page number: 950+. © 1999 College of William and Mary, Marshall Wythe School of Law. COPYRIGHT 2013 Gale Group.

This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.