University of Colorado Law Review > Printed > Volume 91 > Issue 4 > A Short History of the Choice-of-Law Clause

A Short History of the Choice-of-Law Clause

In the field of conflict of laws, private actors are generally granted the power to choose the law to govern their contracts. This is the doctrine of party autonomy. In recent years, this doctrine has been the subject of several excellent histories that draw upon judicial opinions, scholarly writings, and legislative enactments to chronicle changing attitudes toward party autonomy over time. A moment’s reflection, however, reveals that judges, scholars, and legislatures are not the most important actors in this story. The true protagonists are the contracting parties who write choice-of-law clauses into their agreements, without which there would be no need for any doctrine of party autonomy. These drafters and their creations are, however, almost entirely absent from the existing histories.

This Article seeks to remedy this deficit. It provides answers to certain basic questions about choice-of-law clauses that cannot be found in the existing literature. When did they first appear? Have they always been popular? Has the manner in which they are drafted changed over time? It describes how these provisions were first used in the years immediately after the Civil War by companies operating in a small number of industries. It shows how they slowly found their way into a growing number of agreements in the early twentieth century before enjoying a “breakthrough” moment in the early 1960s. And it recounts how contract drafters in the late twentieth century began experimenting with new language that simultaneously expanded the reach of these clauses and prompted the courts to devise new interpretive rules.

This historical account, while interesting in its own right, also has broader implications. First, it underscores the extent to which contract drafters do not always function as rational actors. In some cases, drafters added language to their clauses that was arguably unnecessary. In others, they declined to add language that would have advanced their interests more effectively. Second, it shows that the pattern of contractual change over time in the context of choice-of-law clauses is different than the pattern observed with respect to other types of contractual provisions. This finding suggests the need for new models of contract innovation.

Introduction

When a lawsuit has a connection to more than one jurisdiction, a judge called upon to resolve the dispute must perform a conflict-of-laws analysis to determine which jurisdiction’s law should be applied. In theory, this analysis is straightforward. The judge reviews the facts of the case, determines which state has the closest connection to the parties and the transaction, and applies the law of that state.[2] In practice, things are rarely so neat. Practitioners have long bemoaned the fact that case outcomes in this area can be difficult to predict.[3] And legal scholars have long argued that judges prefer to apply the (more familiar) law of their home jurisdictions and will sometimes use the flexibility inherent in conflicts doctrine to achieve this goal.[4]

Party autonomy—the idea that the parties to a contract may select in advance the law that will govern their contract—offers a solution to both of these problems.[5] When a contract contains a choice-of-law clause, it is easier to predict the outcome of a conflicts analysis because the court will typically apply the law chosen by the parties. The use of such clauses also curtails the ability of judges to engineer the selection of the law of their home jurisdictions. Perhaps for these reasons, choice-of-law clauses have become much more common over time. One recent study found that 75 percent of material contracts executed by public companies contain such a clause.[6] Having obtained the autonomy to select the law to govern their dispute, in short, it appears that many companies are taking full advantage.

The popularity of choice-of-law clauses in contemporary contract practice raises several questions. When did these clauses first appear? Have they always been popular? Has the manner in which they are drafted changed over time? Surprisingly, the existing literature provides few answers to these questions.[7] While the literature contains a number of excellent histories devoted to the concept of party autonomy, these works generally focus on judicial decisions, legislative acts, and scholarly writings.[8] They devote virtually no attention to the history of the contractual instruments by which contracting parties wield their autonomy.

This Article aspires to provide such a history. It shows that the doctrine of party autonomy rests on a contractual foundation. At step one, the parties draft a contract that contains a choice-of-law clause. At step two, the courts interpret the clause and determine whether it is enforceable. At step three, a legislature decides whether to intervene to codify or overturn the judicial decision. At step four, legal scholars discuss and evaluate these judicial decisions and legislative actions. And then the entire cycle begins again as contract drafters draft new choice-of-law clauses that take the foregoing developments into account. Absent the choice-of-law clause, there is no need for any doctrine of party autonomy.[9] And yet contemporary histories of the doctrine generally ignore these contractual provisions.

This oversight is likely attributable to the availability of sources. The doctrinal history of party autonomy is one that can be told through case decisions, scholarly treatises, and legislative enactments. All of these sources can be found on the shelves of a well-resourced law library. The history of the choice-of-law clause, by comparison, is a history that can only be told through a careful review of actual contracts. These agreements are not readily accessible to scholars. They are tucked away in aging file cabinets or stored in boxes in dusty warehouses. In order to study such clauses, one must be able to observe them. And the overwhelming majority of private agreements from long ago cannot easily be observed.

In an attempt to overcome these challenges—and to integrate the choice-of-law clause into the broader history of party autonomy—this Article looks to two original sources.[10] First, it draws upon more than two dozen contract form books published between 1860 and 2019 that contain thousands of “off-the-rack” contracts that were and are used by practicing lawyers to draft agreements for their clients. Second, it draws upon an original, hand-collected data set of more than three thousand choice-of-law clauses that appear in published cases decided between 1900 and 2000. This data set, which was assembled over a period of three years by a team of more than a dozen research assistants, represents the single largest historical collection of choice-of-law clauses currently in existence. Collectively, these sources reveal—for the first time—how contract drafters responded to the multitude of cases, legislative enactments, and treatises that grappled with the challenges posed by choice-of-law clauses beginning in the nineteenth century. In so doing, they make it possible to write the first meaningful history of the choice-of-law clause.

To illustrate the potential utility of such a history, consider the First Restatement of Conflict of Laws, published in 1934. This work took the position that a contract should be governed only by the law of the place of performance or the place of making. It made no reference to the concept of party autonomy and, in so doing, implicitly rejected that concept. The Second Restatement of Conflict of Laws, published in 1971, cast aside the position taken by the First Restatement and expressly endorsed the concept of party autonomy. As a matter of intellectual history, the shift from the First Restatement to the Second Restatement was hugely consequential. As a matter of drafting practice, however, nobody has the slightest idea whether the position taken by the First Restatement had any impact on the willingness of private actors to write choice-of-law clauses into their contracts between 1934 and 1971. It is possible that this seminal moment in the intellectual history of party autonomy had a major impact on drafting practice. It is also possible that it had no impact at all. We simply don’t know.

This Article seeks to answer this question (among others) by looking at choice-of-law clauses through the lens of contract rather than the lens of conflict of laws. This does not mean that conflict of laws is irrelevant to the present inquiry. It is impossible to understand many of the changes to the choice-of-law clause over the past century without taking conflicts doctrine into account. It is merely to point out that the clause itself—which is distinct from the concept of party autonomy—is a worthy object of study that has, at least to date, attracted little attention from legal scholars.

This Article is divided into three parts. Part I focuses on the prevalence of choice-of-law clauses and the debates among judges, legislators, and scholars as to whether these provisions are enforceable. It identifies the first express choice-of-law clauses in the late nineteenth century and traces the slow process by which these clauses came to be used in certain agreements in the years following the Civil War. It explains how these clauses forced the courts to develop an entirely new set of doctrinal rules to determine when such clauses should and should not be given effect. These judicial decisions, in turn, led to these clauses being written into new types of contracts, which prompted an academic debate about the underlying merits of party autonomy as a legal concept. The resolution of this debate in favor of party autonomy led to more choice-of-law clauses and, ultimately, legislative intervention. Part I concludes by discussing the role that these clauses play in modern contract practice.

Part II focuses on the interplay between contract drafters and the courts with respect to how to draft and interpret choice-of-law clauses. In some respects, the language in the standard choice-of-law clause has evolved a great deal over the past 150 years. In other respects, it has remained surprisingly static. The Part shows how each and every word in the modern choice-of-law clause has a specific meaning and, importantly, that this meaning may not be immediately obvious to the casual reader. It then tracks how certain words and phrases that commonly appear in these clauses have changed—or not—in response to judicial decisions over time.

Part III considers the implications of the foregoing analysis for contract law generally. It first argues that the history of the choice-of-law clause provides good reason to question whether contract drafters are, in fact, the rational actors assumed by contract theory. On a number of occasions, drafters have added language to these clauses that is arguably unnecessary. On other occasions, they have declined to add language that would arguably further their interests. The Part then argues that the leading explanation for how and why contract language evolves over time is a poor fit when applied to the choice-of-law clause. It can take decades for innovations to percolate across the contract landscape when it comes to choice-of-law clauses. This slow pattern of change stands in stark contrast to prior studies suggesting that contractual innovation is characterized by rapid and convulsive change. This difference suggests that the process of contractual evolution may vary depending on the precise nature of the provision at issue.

Prevalence and Enforcement

This Part discusses the origins of the choice-of-law clause. It traces the clause’s slow rise to prominence in the nineteenth and twentieth centuries and shows how its growing popularity prompted spirited discussions among judges, scholars, and legislators as to whether and to what extent such provisions should be enforceable. Although the intensity of these debates may seem curious to a modern reader, the question of whether contracting parties had the power to choose the law to govern their agreements was hotly debated in the first half of the twentieth century. For these debates to occur, contracting parties had to start writing choice-of-law clauses into their agreements. This Part recounts the story of that first clause and everything that came after.

Party Intent and Implied Choice-of-Law

In the nineteenth century, courts in the United States generally applied one of two tests to determine the law that would govern a contract.[11] The first test inquired into where the contract was made.[12] The second test inquired into where the contract was to be performed.[13] On occasion, however, the courts would nod to a third test. This test focused on the intent of the parties rather than the place where the contract was made or the place of performance.[14] Today, the notion that party intent should determine what law to apply is generally known as the doctrine of party autonomy.

The origins of party autonomy in Anglo-American law can be traced back to 1760. In that year, Lord Mansfield observed in the English case of Robinson v. Bland that “[t]he law of the place can never be the rule, where the transaction is entered into with an express view to the law of another country, as the rule by which it is to be governed.”[15] This position was subsequently adopted by the U.S. Supreme Court. In 1825, the Court stated in Wayman v. Southard that “universal law [recognizes] the principle that, in every forum, a contract is governed by the law with a view to which it was made.”[16] While express choice-of-law clauses were still largely unknown in this era, the notion that the courts should consider the implied intent of the parties in deciding which law to apply in a given dispute was not.[17]

The earliest reported choice-of-law clauses in the United States were, oddly, interest rate provisions.[18] Consider Butters v. Olds, a case decided in 1860. In this case, the Iowa Supreme Court was called upon to determine the law that would govern a loan agreement. Although the agreement was made in Iowa, the interest was to be paid to the lender in New York. The legal rate of interest in New York was 7 percent. The legal rate of interest in Iowa was 10 percent. The contract called for the borrower to pay interest at a rate of 10 percent. If the contract was governed by the law of the place of performance (New York), it would be invalid for failure to comply with that state’s usury laws. If the contract was governed by the place where the contract was made (Iowa), it would be valid and enforceable.

In concluding that the loan agreement was governed by the laws of Iowa, the Iowa Supreme Court held that the parties had, by selecting the rate of interest permitted under Iowa law, chosen that law to govern their agreement. In its words:

If the rate of interest in the place of the contract differs from that in the State where it is to be performed, it is competent for the parties to stipulate for the rate of interest in either locality, and thus by their own contract determine which law shall govern this incident thereof. In this case the parties did expressly stipulate by their contract for the rate of interest allowed by our statute . . . .[19]

The interest rate provisions in this and similar contracts laid the foundation for the modern choice-of-law clause. In these cases, the courts relied on specific contract language—rather than a holistic analysis of implied intent based on the surrounding circumstances—to hold that the contract would be governed by the law of a particular jurisdiction. While this analytical approach was initially used with respect to a particular class of agreements (loan agreements) and a particular body of law (usury laws), it paved the way for creation of the express choice-of-law clause. Once the principle that the parties could choose a particular state’s usury laws by writing a specific interest rate into their contract was established, it logically followed they should also be able to choose the contract law of that same jurisdiction by including a statement to that effect in their agreement.

The Express Choice-of-Law Clause

The earliest known express choice-of-law clause in the United States appears in a lending agreement executed in 1869.[20] That clause provided that the contract was “made under, and is in all respects to be construed, by the laws of the state of Illinois.”[21] In the decades that followed the Civil War, choice-of-law clauses were gradually introduced into standard-form contracts in industries that did extensive business across state lines.

As life insurance companies began offering policies on a nationwide basis in the 1880s and 1890s, they increasingly utilized choice-of-law clauses to mitigate the legal risks associated with doing business in many different states. In 1882, the Union Central Life Insurance Company entered into an insurance contract with a Virginia resident which stipulated that it was to “be held and construed to have been made in the city of Cincinnati, Ohio.”[22] In 1887, the New York Life Insurance Company issued a policy to a Minnesota resident which stated that “the entire contract contained in the said policy and in this application . . . shall be construed and interpreted . . . according to the laws of the State of New York.”[23] And in 1892, the Mutual Benefit Life Insurance Company issued a policy to an Iowa resident which stipulated that it “shall at all times and places be held and construed to have been made in the city of Newark, New Jersey.”[24] In choosing the law of its home jurisdiction to govern these agreements, each insurer sought to ensure that its contracts would be governed by a single, uniform law, even as it sold policies in many different states.

International transportation companies, which were dominated by foreign interests throughout this era, were also early adopters of express choice-of-law clauses. In 1885, an English company entered into a shipping contract with the Baltimore-based agent of a Chicago businessman to transport cattle from Baltimore to Liverpool. This contract stated that “[a]ny questions arising under this contract or the bill of lading against the steamer or her owners shall be determined by English law in England.”[25] In 1897, a Belgian steamship company sold a ticket to two passengers who wished to travel from Antwerp to New York City.[26] This ticket provided that “all questions arising hereunder are to be settled according to the Belgium law, with reference to which this contract is made.”[27] The purpose of such clauses, again, was to minimize the legal uncertainty that invariably accompanies the act of transporting goods and people from one nation to another.

Late nineteenth-century mortgage lenders doing business across state lines also wrote choice-of-law clauses into many of their contracts. In 1881, for example, the New England Mortgage Security Company entered into a mortgage agreement with an Oregon borrower which was to “be construed according to the laws of Oregon, where the same is made.”[28] In 1886, a Missouri borrower entered into a loan agreement with a New York lender stipulating that “this bond and the interest notes hereto annexed are made and executed under and are in all respects to be governed and construed by the laws of the state of Missouri.”[29] And in 1893, a Minnesota borrower issued a promissory note to a Connecticut lender that contained the following clause: “It is expressly agreed and declared that these notes are made and executed under, and are in all respects to be construed by, the laws of the state of Minnesota.”[30]

It is something of a mystery as to why mortgage lenders typically chose the law of the borrower’s jurisdiction as opposed to the law of the lender’s jurisdiction. In all likelihood, the lenders believed that any litigation arising out of the transaction would occur in the borrower’s home jurisdiction. The odds that the borrower would default, after all, were far greater than the odds that the lender would refuse to make the loan, and it was unlikely that the borrower would be subject to personal jurisdiction outside of his home jurisdiction. Once litigation began, moreover, the courts in the borrower’s home jurisdiction were likely to apply the law of that jurisdiction because that was where the real property was located. If this hypothesis is correct, then the purpose of the choice-of-law clause in mortgage lending agreements was not to give the lender an advantage in litigation or even to ensure a uniform law would be applied to lending agreements concluded in many different states. It was to reduce costs by eliminating any uncertainty as to the governing law at the time of litigation.

In any event, the historical record is clear that express choice-of-law clauses were regularly written into life insurance policies, transportation contracts, and mortgage agreements at the close of the nineteenth century. Inevitably, some of these clauses wound up in litigation before the courts. The reaction of those courts is described below.

The Judicial Response

As express choice-of-law clauses proliferated, the courts were increasingly called upon to determine whether and to what extent such clauses should be enforced. Citing Joseph Story’s acclaimed treatise on conflict of laws, Phillimore’s commentaries on international law, and the decision by the U.S. Supreme Court in Wayman v. Southard—all of which endorsed the general principle of party autonomy—U.S. courts in the late nineteenth century frequently concluded that these clauses were enforceable.[31] In a series of cases in the 1890s, the state supreme courts in Arkansas, Georgia, Minnesota, New York, and Virginia were presented with choice-of-law clauses that sought to displace traditional conflicts rules relating to contracts.[32] In each instance, the court concluded that the parties’ intent—as expressed in their choice-of-law clause—should determine the governing law.

This did not mean, of course, that the courts enforced every choice-of-law clause that came before them.[33] In 1882, a federal court in Missouri was called upon to decide whether a choice-of-law clause selecting the laws of New York in a life insurance contract should be given effect to the exclusion of a Missouri statute designed to protect local policyholders. In concluding that the choice-of-law clause was not enforceable, the court stated:

The statutes of Missouri, for salutary reasons, permit foreign corporations to do business in the state on prescribed conditions. If, despite such conditions, they can by the insertion of clauses in their policy withdraw themselves from the limitations of the Missouri statutes, while obtaining all the advantages of its license, then a foreign corporation can by special contract upset the statutes of the state and become exempt from the positive requirements of law. Such a proposition is not to be countenanced. The defendant corporation chose to embark in business within this state under the terms and conditions named in the statute. It could not by paper contrivances, however specious, withdraw itself from the operation of the laws, by the force of which it could alone do business within the state.[34]

In other cases, the courts cited public policy as a rationale for declining to enforce a choice-of-law clause. In 1893, the U.S. District Court for the Southern District of New York was asked to decide whether a clause selecting the law of England in a bill of lading operated to discharge the carrier from responsibilities imposed by U.S. law. In concluding that it could not, the court observed that:

The objections to the validity of stipulations exempting common carriers from responsibility for negligence, namely, the policy of the law of this country, the unequal situation of the parties, and the lack of sufficient evidence of actual intention and freedom of contract, apply precisely the same to a stipulation for the adoption of the law of another country, as to the original exemption. That [choice-of-law clause selecting English law] is plainly nothing but a further device to secure the same unlawful exemption as the preceding exemption clause, which could not stand alone.

Such an additional stipulation, so far as it relates to the same exemption of liability for negligence, must fall with the latter. Nor can a rule of law founded on public policy be set aside in our own court by any stipulation to adopt the law of another country.[35]

In summary, while many U.S. courts in the late nineteenth century endorsed the notion that parties could select the law to govern their contracts, they also recognized that this autonomy was subject to limits. In particular, these courts took the position that parties could not use a choice-of-law clause to evade the mandatory law of a jurisdiction with a close connection to the parties. They also held that these clauses were unenforceable when contrary to public policy. These limitations notwithstanding, the courts were willing to give effect to these clauses in at least some cases. These decisions, unsurprisingly, led to choice-of-law clauses being written into new types of contracts.

Expansion into New Types of Agreements

In the wake of judicial decisions recognizing choice-of-law clauses as generally enforceable, these clauses gradually came to be incorporated into an ever-wider range of agreements. In 1914, an English company entered into an agreement to acquire substantially all the assets of a company headquartered in Washington state.[36] In 1917, an Illinois-based manufacturer entered into an agreement with an Arkansas-based dealer to sell motor trucks.[37] In 1918, a Chicago-based company entered into a contract to sell heavy machinery to an Indiana-based company.[38] Although each case involved a different type of contract—an asset purchase agreement, an exclusive distribution agreement, and a sales agreement, respectively—each of these contracts contained a choice-of-law clause.

These clauses also began to find their way into employment agreements, particularly when the employee’s work required extensive travel. In 1918, a circus company and one of its employees entered into a contract in which each party acknowledged that the employee’s “performance under this contract shall embrace services and travel in and through the several and various states of the U.S.A. and the Dominion of Canada.”[39] In light of this fact, the parties agreed “that the place of the contract and release, its status or forum, is the District of Columbia” and that “all matters whether sounding or in tort, relating to its validity, construction, and interpretation [shall] be determined to the same extent as if its execution, performance, or cause of action thereon or growing out of the same, actually took place or arose in said District of Columbia.”[40]

The burgeoning motion-picture industry also made extensive use of choice-of-law clauses.[41] In the early 1920s, motion-picture distributors based in New York leased motion pictures to exhibitors across the United States. These agreements frequently contained a choice-of-law clause selecting New York law to govern the provisions of the contract relating to arbitration.[42]

Choice-of-law clauses also appeared in a number of noncommercial agreements during this era.[43] In 1912, a husband living in Virginia sought a divorce from his wife—who was then living in New York—on grounds of desertion.[44] In connection with the divorce proceeding, the parties entered into a separation agreement which stated that it “should be construed according to the laws of Virginia.”[45] In 1914, a wife sought to obtain a divorce from her husband in New York on the grounds of adultery.[46] In connection with this proceeding, the parties entered into an agreement that addressed such questions as alimony, counsel fees, and custody of the couple’s only child.[47] This contract stipulated that it was “to be construed according to the law of the State of New York.”[48]

For ease of reference, the earliest example of a published case referencing a choice-of-law clause in a particular type of contract is set forth in Table 1. This information was gleaned from reported cases rather than actual contracts. Accordingly, it is possible—indeed, it is likely—that the first actual use of a choice-of-law clause in each type of agreement was earlier than the one listed below. Nevertheless, the Table provides a general sense for the slow process by which choice-of-law clauses found their way into different types of agreements before and after the turn of the twentieth century.

This expansion notwithstanding, it is important to emphasize that choice-of-law clauses were not a standard feature of most contracts in the first decades of the twentieth century.[49] Although some companies made regular use of these clauses in certain contexts, many others did not.[50] The cases discussing those clauses that did exist, however, soon began to attract attention from legal scholars. One of these scholars was Joseph Beale.

Table 1. Earliest Choice-of-Law Clause Referenced in Reported Case, by Contract Type.
Contract Type Year of Contract Contract Type Year of Contract
Bond[51] 1869 Distribution Agreement[52] 1910
Mortgage[53] 1872 Asset Purchase Agreement[54] 1914
Prenuptial Agreement[55] 1874 Motion Picture Licensing Agreement[56] 1917
Life Insurance Policy[57] 1880 Sales Agreement[58] 1918
Bill of Lading[59] 1885 Stock Purchase Agreement[60] 1918
Steamship Ticket[61] 1897 Employment Agreement[62] 1918
Trust[63] 1904 Contract of Bailment[64] 1922
Separation Agreement[65] 1909 Rental Agreement[66] 1930

Joseph Beale and the First Restatement

Joseph Beale was a professor at Harvard Law School, the founding dean of the University of Chicago Law School, the reporter for the First Restatement, and the author of an acclaimed multivolume treatise on conflict of laws.[67] In the early decades of the twentieth century, Beale used his position as the leading conflicts scholar in the United States to wage a war against party autonomy and choice-of-law clauses. It was a war that he ultimately lost. In order to understand why, it is important to discuss Beale’s reasons for waging the war in the first place.

Beale’s opposition to the principle of party autonomy stemmed from three sources. First, he believed that the idea that the parties should be permitted to choose the law to govern their agreement was “alien” to the common law tradition.[68] While the continental states of Europe had long endorsed the concept of party autonomy, he argued, it had never been fully embraced by the Anglo-American legal system. In support of this argument, he pointed out that Lord Mansfield had relied upon continental writers when he endorsed the concept in 1760.[69] This was, in Beale’s view, grounds for concern: “Is it then permissible for us to base principles of the Conflict of Laws on civil-law authorities? It is submitted in general that this should not be done.”[70]

Second, Beale argued that the U.S. Supreme Court had been less than steadfast in its commitment to the principle of party autonomy in the years since the Court first endorsed it in 1825.[71] In support of this argument, he identified a number of more recent conflicts cases in which the Court had articulated a conflicts rule relating to contracts—to apply the law of the place where the contract was made, for example—without making any mention of party intent. The persuasiveness of Beale’s analysis in this regard is undercut by the fact that the issue of party intent was apparently not raised as an issue in many of the cases that he cited. Nevertheless, he invoked the Court’s failure to mention party intent in every conflicts case that came before it as proof that the Court was growing increasingly hostile to the principle of party autonomy.[72]

Third, Beale argued that party autonomy was inconsistent with the vested rights theory of conflict of laws. This theory posited that all legal rights vested in a specific time and place and thereafter traveled with the plaintiff wherever he went.[73] For example, if a person were injured in an accident in Mississippi and subsequently brought a lawsuit against the tortfeasor in Alabama, the suit would be governed by Mississippi law because the right to bring a tort claim had “vested” in Mississippi at the moment of the injury.[74] The proponents of the vested rights theory believed that each state was sovereign within its own borders and that no state could make a law that would operate in any other state.[75] It logically followed from this proposition that the parties to a contract were powerless to select the law that would govern their agreements. As Beale explained:

The fundamental objection to [party autonomy] in point of theory is that it involves permission to the parties to do a legislative act. It practically makes a legislative body of any two persons who choose to get together and contract . . . and gives to the parties what is in truth the power of legislation so far as their agreement is concerned.[76]

If all lawmaking power was vested in the state, and if each state lacked the power to legislate outside of its territorial borders, then it was irrelevant whether a contract contained an express choice-of-law clause. The only legal rights that could possibly have been invoked by a plaintiff were those that had been vested by the state with sovereign control over the territory where the cause of action arose.

If Beale had simply made these arguments in the context of his own academic writings, they would have been influential enough; Beale towered over the field of conflicts law for more than fifty years.[77] As mentioned above, however, Beale also served as the reporter for the First Restatement of Conflict of Laws. In this capacity, he was tasked with “restating” the common law of conflicts. When it came time to draft those sections of the First Restatement dealing with the issue of party autonomy, Beale put forward his own vision of what the law should be rather than restating the law as it was.[78] Notwithstanding the extensive caselaw endorsing the principle of party autonomy discussed above, the First Restatement contains not a single reference to it. Indeed, the First Restatement is devoid of any hint of the possibility that party intent has a role to play in conflict of laws.[79]

Shortly after the First Restatement was published in 1934, Beale made an observation and a prediction. He observed that the “tendency [among the courts] is to abandon the law intended by the parties as the law to govern the validity of a contract.”[80] He then predicted that it was “probable that before many years have passed the influence of the American Law Institute will have led to the abandonment of the doctrine of intention of the parties.”[81] As it turns out, Beale was incorrect on both counts.

In an article published in 1942, Arthur Nussbaum took issue with Beale’s “observation” that U.S. courts were becoming less and less likely to enforce choice-of-law clauses.[82] Nussbaum noted that the intent theory was one of the most “time-honored” and “universally adopted” rules of private international law.[83] He observed that the U.S. Supreme Court had endorsed the general principle of party autonomy on numerous occasions and had commented favorably on the use of choice-of-law clauses. Nussbaum also identified a number of methodological problems with respect to Beale’s case analysis purporting to show a trend away from enforcing choice-of-law clauses.[84] After canvassing the cases that Beale had relied upon to support his observation, Nussbaum concluded that they provided little to no support for the empirical claim that U.S. courts were becoming less likely to enforce choice-of-law clauses.[85] Beale’s descriptive account of recent judicial practice, in Nussbaum’s assessment, was an extended exercise in wishful thinking.

Beale’s prediction as to the future also proved inaccurate.[86] In 1937, just three years after the publication of the First Restatement, the U.S. Supreme Court doubled down on its prior decision in Wayman v. Southard when it observed that a choice-of-law clause in a contract between a Pennsylvania oil company and a Connecticut insurer was “undoubtedly” enforceable because “[i]n every forum a contract is governed by the law with a view to which it was made.”[87] In 1946, a panel of judges sitting on the Second Circuit made clear that they were familiar with the First Restatement and nevertheless believed that the parties should have the autonomy to select the law to govern their agreement:

Under orthodox conflict of laws rules this provision [for rescission] is to be given effect by the forum according to the law of the place of performance. Restatement, Conflict of Laws (1934) Secs. 358, 359. Where, however, there is no single place of performance, and there are many jurisdictions with which the contract has close association, there seems no reason, apart from the policy of the forum, why the parties cannot specify the law of one jurisdiction as controlling, so long as there is that sufficient relationship to make it reasonable that the law chosen should apply. Decisions to the seeming contrary have been so on the basis of the policy of the forum or on the ground that there was an apparent place of performance the law of which forbade choice of another law. Since no fixed policy to the contrary has been established in this court, [the choice-of-law clause selecting New York law] is controlling and the rights of the parties herein are to be determined by the law of New York.[88]

The final blow came in 1952, when the Uniform Commercial Code (UCC) was published.[89] The drafters of the UCC included a provision that expressly permitted the contracting parties to choose the law of any state that bore a “reasonable relation” to the contract.[90] At this point, the intellectual debate shifted away from whether the intent of the parties should be considered to whether party autonomy should be subject to any limits. This development, unsurprisingly, led to yet another uptick in the number of contracts that contained choice-of-law clauses, as recounted below.

From Novelty to Mainstream

It is difficult to determine the impact of the foregoing developments—the publication of the First Restatement, the ensuing academic debates about the nature and scope of party autonomy, and the publication of the UCC—on the prevalence of choice-of-law clauses in U.S. contracts. This is because there is no easy way to estimate, as a historical matter, what percentage of contracts contained a choice-of-law clause in the early to mid-twentieth century. To calculate this number accurately, it is necessary to determine, first, the total number of contracts executed in a given year and, second, how many of these contracts contained choice-of-law clauses. From the vantage point of 2020, it is simply not possible to gather this information.

It is possible, however, to obtain a rough sense for the prevalence of such clauses by looking to books of form contracts that were published during the relevant time periods.[91] In an era before photocopiers—let alone computers and word processors—lawyers would routinely consult form books containing samples of many different types of contracts when called upon to draft a particular type of agreement.[92] General form books typically contained hundreds of agreements, organized by type, that could be quickly and cost-effectively deployed by the contract drafter when the need arose.[93] Since these books provide a historical record of what provisions were typically included for specific types of agreements at a particular point in time, they offer a means by which scholars can get a general sense for the prevalence of the choice-of-law clause in a particular era. One need only select a well-known form book from a given year, count the number of contracts in the book, and calculate the percentage of those contracts containing choice-of-law clauses. As a judge on the Pennsylvania Court of Common Pleas put it in 1930: “An examination of the form books, while not being recognized as stating the law, gives some indication of what the general practice is.”[94]

With this insight in mind, I reviewed twenty-five form books with the aid of two research assistants. The earliest form book dated to 1860. The contracts in that book contained not a single choice-of-law clause. The most recent form book dated to 2019. Sixty-nine percent of the contracts in that book contained a choice-of-law clause. The bulk of our time and attention was spent on form books published between these years. With respect to each book, we recorded the total number of contracts contained therein as well as the number of those contracts that contained a choice-of-law clause.

There are, of course, methodological problems with using the form books in this way. On the one hand, form books could be a leading indicator of actual party practice. If this is the case, then changes in the form books may precede changes in real-world contract drafting. On the other hand, form books could also be a trailing indicator of actual party practice. If this is the case, then changes in the form books may lag behind changes in real-world contract drafting. It is also possible that the form books represent a stylized version of contract practice that in no way reflects actual contracting practice during the era. If this is the case, then the form books may present a portrait of contract practice that affirmatively misleads the reader. Nevertheless, the form books arguably represent the best sources available to a researcher trying to ascertain the frequency with which choice-of-law clauses were incorporated into contracts drafted more than a century ago.

While looking to form books is an act of desperation, there is some reason to believe that these books offer a reasonably good guide to actual contract practice when it comes to choice-of-law clauses. A review of a modern form book available through Lexis in 2019—Forms from Rabkin & Johnson, Current Legal Forms with Tax Analysis—turned up 1,381 contracts. Nine hundred fifty-two of these agreements—69 percent—contained choice-of-law clauses. A recent empirical study that reviewed half a million actual contracts filed with the U.S. Securities and Exchange Commission (SEC) between 2000 and 2016 found that 75 percent of these agreements contained choice-of-law clauses.[95] The fact that these two percentages are in the same general ballpark—69 percent versus 75 percent—provides some comfort that form books from the past might offer a reasonably accurate measure of the prevalence of choice-of-law clauses in earlier eras.

A review of contract form books published between 1910 and 1965 suggests that although the choice-of-law clause was widely adopted by companies in a few select industries early on, such clauses were not particularly common prior to the early 1960s. Consider Sayler’s American Form Book, published in 1913, and Tiffany’s Legal and Business Forms, published in 1915.[96] Together, these two books contain 471 form agreements. Only two of these agreements include a choice-of-law clause (0.4 percent of the total). A review of three form books published in the 1920s tells a similar story. Collectively, these form books contained 901 form agreements.[97] Only thirteen of these agreements contained choice-of-law clauses (1.5 percent). In 1938, West published a massive, multivolume collection titled Modern Legal Forms that purported to reflect current trends in contract drafting.[98] A review of two volumes from this collection turned up 263 contracts but only six choice-of-law clauses (2.3 percent). Looking ahead to the 1940s, a form book published in 1946—Gordon’s Modern Annotated Forms of Agreement—contained 431 contracts but only eleven choice-of-law clauses (2.6 percent).[99] The 1946 edition of Jones Legal Forms Annotated tells a similar story.[100] Although it contained 678 contracts, only three of those contracts contained choice-of-law clauses (0.4 percent).

The fact that choice-of-law clauses were rarely written into contracts in form books published in the 1920s, 1930s, and 1940s suggests that Beale’s decision to omit any discussion of party autonomy from the First Restatement was of relatively little consequence from a drafting perspective. While it is possible that the position laid down by the First Restatement discouraged parties who would otherwise have written such clauses into their agreements from doing so—an empirical question that is extremely difficult to answer—it does not appear that this work prompted contract drafters to remove existing provisions from their agreements. The evidence from the form books suggests that, in most cases, there was nothing to remove.[101]

Not until the 1950s did drafting practice undergo a meaningful shift. While virtually all of the contracts in the fourteen-volume American Jurisprudence Legal Forms Annotated (1953) lacked choice-of-law clauses, the inside front cover of each volume contained a “checklist” of things to remember when drafting forms.[102] The third item on the checklist encouraged contract drafters to ask: “Should there be a declaration that the laws of a specific jurisdiction will govern interpretation, validity, performance, etc.?” The form books from the early 1960s are the first to contain a meaningful number of contracts with choice-of-law clauses. Approximately 21 percent of the 140 contracts in Forms of Business Agreements with Tax Ideas (1961) contained a choice-of-law clause.[103] And 19 percent of the 118 contracts in Forms for Commercial Transactions Under the Uniform Commercial Code (1963) contained such a clause. This marked a significant shift in practice, as shown in the table below.[104]

A screenshot of a cell phone Description automatically generated

It appears that the early 1960s represented something of a watershed moment for the choice-of-law clause. Although such clauses had been around for almost a century by this point, and although they had long been embraced by companies in a few select industries, there is no evidence that these clauses were written into contracts outside of these industries with any sort of regularity until the early 1960s. Indeed, I was unable to find a single form book published prior to 1961 where more than 3 percent of the contracts contained choice-of-law clauses. While the clause was known to prior generations of contract drafters, it was not widely used until the early 1960s.

There are many possible explanations for why the choice-of-law clause gained traction at this particular moment in history.[105] Due to space limitations—and the uncertain reliability of the underlying sources—only one possible explanation will be addressed here. This is the possibility that the enactment of the UCC spurred more parties to write choice-of-law clauses into their agreements. As discussed above, the draft UCC contained a provision that specifically directed courts to enforce choice-of-law clauses in commercial contracts when the chosen state bore a “reasonable relation” to the contract. Although the UCC was first published in 1952, it was substantially revised in 1956 and was not enacted by most states until the early 1960s.[106] It may not be a coincidence that there was an uptick in the number of choice-of-law clauses appearing in form books at the same moment when many states were in the process of enacting a statute that specifically authorized courts to enforce choice-of-law clauses in commercial agreements.[107] To be sure, this account is speculative. There is no easy way to determine why a widely dispersed body of private actors began to make changes to their private agreements many decades ago. It is, however, an explanation that is at the very least consistent with the patterns of practice suggested by the form books.

The Second Restatement and State Statutes

The Second Restatement, which was finalized in 1969 and published in 1971, staked out a position on the issue of party autonomy that was very different from that of the First Restatement. With respect to default legal rules, the Second Restatement stipulated that the parties could choose the law of any state on the theory that they could have simply written the rule into their contract.[108] With respect to mandatory legal rules, the Second Restatement took the position that parties could select the law of a particular state to govern their agreement but that this ability was subject to two limits. The first limit requires that either the chosen law have a substantial relationship to the parties or the transaction or there be a reasonable basis for the parties’ choice.[109] The second limit requires that the chosen law not be contrary to a “fundamental policy of a state which has a materially greater interest . . . in the determination of the particular issue and which . . . would be the state of the applicable law in the absence of an effective choice of law by the parties.”[110]

Each of these limitations can be easily traced to prior legislative or judicial practice. The “substantial relationship” requirement bears a close resemblance to the “reasonable relation” requirement that appears in the UCC.[111] The “reasonable basis” requirement was derived from cases upholding the validity of clauses selecting the law of a significant commercial jurisdiction such as England or New York as a “neutral” law.[112] And the “fundamental policy” requirement can be traced back to cases from the late nineteenth century in which U.S. courts refused to uphold choice-of-law clauses in the face of local statutes requiring the application of forum law to certain types of disputes.[113] Unlike the First Restatement, the Second Restatement actually sought to restate the law as it was, as opposed to how its drafters wanted it to be. Perhaps as a result, section 187 is now among the most widely followed provisions in the Second Restatement.[114]

The enforceability of choice-of-law clauses is not, however, always determined by common law rules of the sort laid down by the Second Restatement. As choice-of-law clauses proliferated in the decades after 1960, many state legislatures chose to enact statutes that preempt these common law rules. Some of these enactments directed state courts to apply the law of the forum even when the parties had expressly selected the law of another jurisdiction. These statutes served to restrict party autonomy. Other enactments directed state courts to enforce choice-of-law clauses even when the law chosen lacked a substantial relationship to the parties or the transaction. These laws operated to expand party autonomy. Both types of statutes are discussed below.

Statutes Restricting Party Autonomy

State statutes that direct state courts to apply the law of the forum even when the parties have selected the law of another state have a long history. In the 1890s, Missouri enacted a statute designed to protect its residents against so-called “forfeiture” provisions in life insurance contracts.[115] The statute made clear that such provisions could not be enforced even when the contract in question contained a choice-of-law clause selecting the law of a state other than Missouri.[116] In this way, the state legislature restricted the ability of parties to wield their autonomy to select the law of another state.

In the second half of the twentieth century, states enacted other statutes restricting party autonomy with respect to specific types of contracts. The Uniform Consumer Credit Code, approved in 1968, directs courts to refuse to enforce choice-of-law clauses selecting the law of a state other than that of the borrower’s residence.[117] In the 1970s and 1980s, many states enacted franchise acts that directed state courts not to enforce any contract provision—including a choice-of-law clause—that would amount to a “waiver” of a right guaranteed to the franchisee under the statute.[118] Other states enacted similar anti-waiver provisions with respect to consumer contracts and employment contracts.[119] The overall effect of such provisions was (and is) to limit the ability of lenders, franchisors, manufacturers, and employers to utilize choice-of-law clauses to select the law that will govern their contractual relationships with borrowers, franchisees, consumers, and employees.[120]

While the foregoing statutes restricted party autonomy to address concerns about unequal bargaining power, other statutes restricted party autonomy based on a different set of concerns. In the first two decades of the twenty-first century, a number of state legislatures passed laws declaring choice-of-law clauses selecting non-U.S. law to be unenforceable under certain circumstances. The principal target of such statutes was Islamic law (or sharia), which was perceived by elected representatives in these states to pose a threat to American values.[121] The statutes in question generally direct state courts not to give effect to choice-of-law clauses selecting foreign law when to do so would abridge rights conferred upon the state’s residents under the U.S. Constitution or the state constitution.[122] The relevance of these statutes to real-world disputes is unclear—to date, it appears that such laws have actually been used to invalidate a choice-of-law clause selecting foreign law on only one occasion.[123] Nevertheless, such statutes represent a noteworthy legislative intervention designed to limit the autonomy of contracting parties to choose the law to govern their agreement.

Statutes Expanding Party Autonomy

Over the past few decades, state legislatures in the United States have also enacted several pieces of legislation designed to expand the autonomy of contracting parties. Under the common law approach set forth in section 187 of the Second Restatement, the parties’ choice of law will generally be honored if the chosen law has a “substantial relationship to the parties or the transaction” or if there is a “reasonable basis” for their choice. In the 1980s, however, a number of states enacted statutes that relaxed these restrictions for high-dollar-value commercial contracts. In 1984, for example, New York enacted a statute directing its courts to enforce choice-of-law clauses selecting New York law in commercial contracts for more than $250,000 even when the parties and the transaction lacked a “reasonable relation” to New York.[124] The legislature was transparent about its motivation in passing this law—it hoped to divert legal business to New York and away from other jurisdictions, thereby generating more business for New York lawyers.[125] The practical effect of this statute was to encourage companies with no other connection to New York to select that state’s law to govern their agreements, without any concern that the choice-of-law clause would be invalidated for the lack of any “substantial relationship” to New York.

In the years that followed, a number of other states followed New York in requiring their courts to enforce choice-of-law clauses selecting their law even where the transaction lacked a substantial relationship to the state.[126] In every case, the statutes did not direct state courts to enforce choice-of-law clauses generally. Instead, the statutes directed state courts to enforce clauses contained in qualifying contracts if, and only if, the clause selected the law of the enacting state. In light of this requirement, it would seem these enactments—like the one pioneered by New York—are less concerned with principles of party autonomy than they are with generating business for in-state lawyers.

A statute enacted by North Carolina in 2017 goes even further. This statute stipulates that a choice-of-law clause selecting North Carolina law in a business contract is enforceable even when the parties and the transaction lack a “reasonable relation” to the state.[127] The statute then goes on to provide that the same result should be obtained even when the contract contained a provision that was “contrary to the fundamental policy of the jurisdiction whose law would apply in the absence of the parties’ choice of North Carolina law.”[128] The end result is a legal regime in which the North Carolina courts will apply that state’s law to any business contract selecting the law of North Carolina, even when the transaction lacks a reasonable relation to the state and even when its law is contrary to a fundamental policy of a jurisdiction with a closer connection to the dispute.

  1. The Choice-of-Law Clause in the Twenty-First Century

Today, choice-of-law clauses are everywhere. A recent study of every contract filed with the SEC between 1996 and 2012 found that 70 percent contained a choice-of-law clause.[129] A subsequent study looking at the same database covering the time period from 2000 to 2016 pegged the number at 75 percent.[130] When the focus is narrowed to international contracts—agreements between a U.S. company and a foreign counterparty—the percentage rises still higher. One study of international supply agreements filed with the SEC between 2011 and 2015 found that 99 percent contained a choice-of-law clause.[131] There is little question that these clauses have become standard features in commercial agreements over the past century.

This outcome represents the culmination of a long conversation between contract drafters on the one hand, and judges, scholars, and legislatures on the other. Contract drafters wrote the first express clauses into their agreements in the nineteenth century, thereby initiating a judicial conversation about the proper role of party autonomy in conflict of laws. Other contract drafters incorporated these clauses into an ever-wider range of agreements in the early twentieth century, triggering the scholarly revolt against party autonomy led by Joseph Beale. Still other contract drafters began to make more extensive use of these clauses in the early 1960s, thereby prompting state legislatures to pass new laws spelling out the proper scope of party autonomy in a range of contexts involving consumers, employees, insureds, and franchisees. The end result is a complex tapestry of contract provisions, common law rules, statutes, and judicial commentary that defines the scope of party autonomy in the modern era.

* * *

The foregoing discussion focused on the prevalence of choice-of-law clauses and the conversations among judges, scholars, and legislatures as to whether and to what extent they should be enforceable. This is not, however, the only topic of historical interest relating to choice-of-law clauses. Another important question is whether the language in these clauses has changed in any meaningful way over the past 150 years. This issue is taken up below.

  1. Drafting and Interpretation

The first known modern choice-of-law clause, which appeared in 1869, provided that the contract was “made under, and is in all respects to be construed, by the laws of the state of Illinois.”[132] In the decades that followed, thousands—and eventually millions—of choice-of-law clauses were written into other legal agreements. Inevitably, the language in these clauses changed and evolved. Certain words and phrases fell out of favor and were replaced. New legal issues arose that necessitated the addition of new language. These changes did not occur overnight. Indeed, the process of change was, in many cases, slow and halting. Over the course of decades, however, the typical choice-of-law clause gradually came to exhibit certain features that were distinct and different from those very first clauses drafted at the close of the nineteenth century.

This Part provides a historical account of these changes. As a prelude to this account, however, it is necessary to discuss questions of research methodology and data collection. It is obviously impossible to review and inspect every choice-of-law clause used in the tens of millions of U.S. contracts that entered into force over the past 150 years. In order to overcome these challenges, I turned to a somewhat unusual source: published cases. Over a period of several years, I worked with more than a dozen research assistants to comb through such cases in search of choice-of-law clauses. Whenever we found a clause referenced in a case, we inputted that clause—along with the year the contract containing the clause was signed and the type of contract at issue—into a spreadsheet. When our work was complete, we had collected 3,104 choice-of-law clauses that selected the law of a U.S. jurisdiction.[133] A comprehensive discussion of the data collection methods used to assemble this data set—along with several caveats to which its use is subject—is set forth in Appendix A. A table listing the number of clauses from various decades is set forth below.

Table 2. Number of Choice-of-Law Clauses Collected from Published Cases, by Contract Decade.
Decade

Contract Signed

Number of

Clauses

1880s 15
1890s 51
1900s 44
1910s 31
1920s 43
1930s 53
1940s 79
1950s 123
1960s 236
1970s 398
1980s 1020
1990s 1011
Total 3104

This Part explores the changing language in choice-of-law clauses across a number of dimensions. First, it looks to see whether certain words and phrases associated with the First Restatement—such as clauses that specifically reference the “place” of the contract—have become more or less common over time. Second, it discusses the relative frequency of the words “interpret” and “construe.” Third, it looks at the changing role of the word “govern.” Fourth, it seeks to determine whether “broad” choice-of-law clauses—those that select the tort and statutory law of the chosen jurisdiction in addition to its contract law—have become more common. Fifth, it discusses the rise of clauses that exclude the conflicts rules of the chosen jurisdiction. Sixth, it looks at whether clauses ever select the procedural law of the chosen jurisdiction. Finally, it briefly discusses additional words and phrases that may be of interest to scholars but that rarely appear in actual choice-of-law clauses.

Use of “Made,” “Executed,” “Performed,” and “Place”

Under the traditional vested-rights approach to conflict of laws, a contract was governed either by the law of the place where it was made or the law of the place where it was to be performed. To determine where a contract was made, the courts would generally look to where the offer was accepted. To determine where a contract was to be performed, the courts would generally look to where the promise was to have been performed. As principles of party autonomy gained traction in conflict of laws, and as companies began drafting contract language to take advantage of this autonomy, the resulting clauses would sometimes make express reference to one or both of these rules.

In 1881, for example, an insurance company issued a life insurance policy stipulating that “[t]his policy is a contract, made and executed in the State of New York, and shall be construed only according to . . . the laws of that State.”[134] In 1902, a different insurance company issued a policy stating that it was “a contract made and to be performed in the state of Wisconsin, and shall be construed only according to . . . the laws of said State.”[135] In 1902, still another insurance company issued a policy providing that “the place of this contract is expressly agreed to be in the city of Binghampton [sic], N. Y.”[136] In each case, the clause in question was drafted in such a way as to preemptively answer the questions that insurance companies expected courts committed to a theory of vested rights to ask—where was the contract made and where was it to be performed?

The vested rights approach to conflict of laws dominated the landscape in the early twentieth century.[137] In the 1950s and 1960s, however, U.S. courts began to experiment with new and different approaches, a process of experimentation now commonly described as the “Conflicts Revolution.”[138] In the realm of contracts, the primary effect of this Revolution was to de-emphasize the place of making and the place of performance as factors in determining the governing law.[139] Instead, courts began to inquire into which jurisdiction had the closest connection to the parties and the transaction.[140] This new approach, which was first adopted by the New York Court of Appeals in the mid-1950s, was eventually incorporated into the Second Restatement.[141] Section 188 of that work provides that, in the absence of a choice-of-law clause, “the rights and duties of the parties with respect to an issue in contract are determined by the local law of the state which, with respect to that issue, has the most significant relationship to the transaction and the parties under the principles stated in § 6.”[142] Under this new test, the place where the contract was made and the place where the contract was to be performed were factors for the court to consider in deciding which law to apply. At the end of the day, however, these factors were now part of a broader holistic analysis keyed to the concept of the most significant relationship.

As the Conflicts Revolution gathered steam, one would perhaps expect to see a decrease in the proportion of choice-of-law clauses that contained words such as “made” or “executed” or “performed” or “place.” All of these words, after all, were originally included in these clauses to answer questions posed by a (now-outdated) system for determining the law that should be applied to a contract with a connection to more than one jurisdiction.[143] There is, however, scant evidence that the Conflicts Revolution had much (if any) impact on the way that choice-of-law clauses were drafted, as shown below.

A close up of a map Description automatically generated

On the one hand, certain words associated with traditional conflicts rules have become slightly less common over time. For example, the proportion of clauses that refer to the place where the contract was “executed” declined from 5 percent in the 1940s to 3 percent in the 1990s. And the proportion of clauses that refer to the “place” of the contract declined from 4 percent in the 1940s to 1 percent in the 1990s. Conversely, other words associated with traditional conflicts rules have become slightly more common over time. The proportion of clauses that refer to the place where the contract was “made” rose from 6 percent in the 1940s to 7 percent in the 1990s. And the proportion of clauses that contain some variant of the word “perform” rose from 10 percent in the 1940s to 12 percent in the 1990s. This suggests that while the Conflicts Revolution of the 1950s and 1960s may have upended choice-of-law doctrine across the United States, its impact on how choice-of-law clauses were drafted was negligible.

Use of “Interpret” and “Construe”

In reviewing many thousands of choice-of-law clauses, two words appear over and over again. These words are “interpret” and “construe.” Some clauses provide that they shall be “interpreted in accordance with the laws” of a particular state. Other clauses provide that they shall be “construed in accordance with the laws” of a particular state. In light of this variation in drafting practice, it is sensible to inquire whether these two words are synonyms. In order to answer this question, it is necessary to take a brief detour into the law of contracts as it existed in the late nineteenth century.

In theory, it is possible to draw a distinction between the act of “interpreting” a contract and the act of “construing” that same agreement. The word “interpretation” technically refers to the process judges use to determine the meaning of the words actually used in an agreement.[144] The word “construction,” by contrast, refers to the process judges use to determine the meaning of the contractual relationship taken as a whole.[145] Under this framework, a contract must be interpreted before it may be construed. If the words in the contract are unambiguous as an interpretive matter, there is no need to construe the contract. If the words are ambiguous, however, the court may then draw upon various canons of construction to help determine their meaning.

While the distinction between interpretation and construction may well be important as a matter of contract theory, courts and commentators long ago concluded that this particular distinction was of little practical significance. As the author of an 1886 treatise on commercial contracts explained:

Throughout the present volume interpretation and construction are used as synonymous words. Attempt, however, has been made to distinguish between them and interpretation employed to denote merely the process of ascertaining the meaning of the terms of the contract, and construction to describe the further process of bringing this meaning into consonance with established rules of law. But they are used interchangeably by the judges, in the digests, and by the large majority of text writers, and notwithstanding the high authority of those who have endeavored to enforce the distinction, no sufficient reason for its adoption is perceived.[146]

Other commentators writing in the late nineteenth and early twentieth centuries took a similar position. One observed that “[t]he terms ‘interpretation of a contract’ and ‘construction of a contract’ are usually used interchangeably.”[147] Another pointed out that

the rules of interpretation and construction are necessarily more or less blended, since, in determining the meaning of a word, we have sometimes to appeal to the context, just as, in determining the meaning of the context, we have to throw ourselves into the position of the author when he used the particular words.[148]

As a matter of contract law, therefore, the distinction between “interpretation” and “construction” was not viewed as particularly meaningful at the turn of the twentieth century.

This was also broadly true as a matter of conflict of laws. As early as 1834, for example, Justice Story wrote that “it is now a principle, generally received, that contracts are to be construed and interpreted according to the laws of the state, in which they are made, unless from their tenor it is perceived, that they were entered into with a view to the laws of some other state.”[149] In 1905, Francis Wharton stated much the same rule in his treatise without drawing any distinction between the two terms.[150] Even Joseph Beale, writing in 1934, used the words “interpret” and “construe” interchangeably when discussing the parties’ ability to select the law that was to govern their contract.[151] In the eyes of most conflict-of-laws commentators in this era, therefore, the words “interpret” and “construe” meant the same thing.[152]

A review of the clauses in the data set reveals that it is far more common for these clauses to utilize the word “construe” than the word “interpret,” as illustrated below.

A screenshot of a map Description automatically generated

Notwithstanding diligent searching, I have not found any cases where a court attached any significance to the parties’ decision to utilize the word “interpreted” rather than the word “construed” in a choice-of-law clause. While it is probably more advisable to use the word “construe” in light of the technical distinctions outlined above, the courts invariably treat the two words as synonymous. Perhaps as a consequence, the relative proportion of clauses using the words “interpret” and “construe” has changed very little over the past few decades.

Use of “Govern”

In deciding what law to apply to resolve a contract dispute, courts have traditionally distinguished between several issues. The first is whether the contract is valid, that is, whether the promises made by the parties in the contract are binding. As a general rule, courts will always apply the law of the forum—rather than any law chosen by the parties—to resolve this question because the issue of validity is a threshold issue that must be addressed to determine whether a contract exists in the first place. The second issue is whether the choice-of-law clause should be given effect. A court may conclude, for example, that the parties entered into a valid contract but that the choice-of-law clause should not be given effect because applying the chosen law would contravene some strong public policy of the forum. Whether a clause should be given effect is, again, an issue generally determined by the law of the forum.

Once a court has determined that the contract is valid and the choice-of-law clause should be enforced, it must then determine the breadth of the choice-of-law clause. Does the clause select all of the contract law of the chosen jurisdiction, including any rules relating to default terms, performance, defenses, breach, and remedies? Or does it only select that portion of the chosen jurisdiction’s contract law that specifically relates to the interpretation or construction of otherwise valid contracts?

In answering this question, some courts have attached great significance to the precise words used by the parties in their choice-of-law clause. These courts have held that the use of the word “govern” signals an intent to select all of the contract law of the chosen jurisdiction. The use of the words “interpret” or “construe,” by contrast, only signals an intent to select that portion of the chosen jurisdiction’s contract law that specifically relates to interpretation.[153] To be clear, most U.S. courts have not endorsed this interpretive rule. The majority of courts—reasoning that it would be curious indeed for the parties to select the interpretive law of the chosen jurisdiction while leaving unanswered what contract law should be applied to other issues—have held that “govern,” “interpret,” and “construe” are synonyms. Beginning in the late 1970s, however, a minority of U.S. courts took the position that the words “interpret” and “construe” signal the parties’ intent to select only that portion of the chosen jurisdiction’s contract law that specifically relates to the interpretation or construction of otherwise valid contracts.[154]

Although only a few case decisions ever formally endorsed this distinction, the relative proportion of choice-of-law clauses utilizing the word “govern” experienced a sharp uptick beginning in the 1970s. Between 1940 and 1970, the percentage of choice-of-law clauses that contained this word held relatively constant at around 40 percent. In the 1970s, however, the percentage jumped to 55 percent. In the 1980s, it surged to 68 percent. In the 1990s, it rose again to 73 percent. Significantly, this uptick occurred even as the prevalence of “interpret” and “construe” remained relatively constant.

 

A screenshot of a cell phone Description automatically generated

This shift in drafting practice cannot be wholly attributed to judicial decisions suggesting that the words “interpret” and “construe” were too narrow to encompass all of the chosen jurisdiction’s contract law. The first judicial decision to assign a narrow meaning to these words dates to 1978, which postdates the initial uptick in choice-of-law clauses containing the word “govern” by several years. It seems likely, however, that these judicial decisions assigning a broader scope to the word “govern” played some role in driving the increase in the proportion of clauses that contain that word in the 1980s and 1990s.

This uptick is particularly noteworthy because the caselaw assigning a narrow meaning to the words “interpret” and “construe” did not represent the majority rule. As discussed above, most U.S. courts that addressed this issue in the late 1970s and early 1980s concluded that “interpret,” “construe,” and “govern” were synonyms.[155] This fact notwithstanding, a significant number of contract drafters apparently chose to rewrite their choice-of-law clauses to include the word “govern” to guard against the risk that a court might assign their clause a narrow meaning.

Noncontractual Law

The discussion above considered whether a particular choice-of-law clause selected all of the chosen jurisdiction’s contract law or only the portion of that jurisdiction’s contract law relating to contract interpretation. This was not, however, the only scope question to come before the courts in the late 1970s. In other cases, the courts were asked to decide whether a given choice-of-law clause merely selected the contract law of the chosen jurisdiction or whether it selected the tort and statutory law of the chosen jurisdiction

In the mid-1970s, for example, an investor entered into a contract with a stockbroker in the city of New Orleans.[156] This contract contained a standard choice-of-law clause stipulating that “[t]his contract shall be governed by the laws of the State of New York.”[157] After the stockbroker lost all of the investor’s money, the investor brought suit against the stockbroker in New York state court for breach of contract, negligence, and fraud. The court held that while the law selected in the contract’s choice-of-law clause covered the breach of contract claim, it did not cover the plaintiff’s claims for negligence and fraud because the language in the clause did not sweep broadly enough to cover noncontractual claims. After performing a conflicts analysis, the court concluded that while the plaintiff’s contract claims were governed by the laws of New York, per the choice-of-law clause, his claims for negligence and fraud were governed by the law of Louisiana, per the court’s conflicts analysis.[158]

In the wake of this and similar decisions, one might expect to see an increase in the proportion of choice-of-law clauses that select the noncontractual law of the chosen jurisdiction. Lawyers tasked with drafting choice-of-law clauses on behalf of their corporate clients, after all, generally wish to reduce legal uncertainty to the greatest possible extent. And broad choice-of-law clauses make it easier to predict the law that will govern any noncontractual claims brought against the company in the future.[159] There is, however, no evidence that contracting parties across the United States responded to these decisions by writing broader choice-of-law clauses into their contracts in the 1980s and 1990s. Throughout this time period, parties continued to draft clauses that selected the contract law—and only the contract law—of the chosen jurisdiction.

To understand the basis for this conclusion, it is necessary first to understand that there are many different ways to draft a clause that selects the non-contract law of a given jurisdiction. A clause could state that the chosen law shall apply to all claims that “relate to” or have some “connection with” the contract.[160] Alternatively, a clause could select the “tort law” of the chosen jurisdiction or state that the “relationship” between the parties shall be governed by the law of a given state.[161] Even taking these different possible formulations into account, however, just 2 percent of the 1970s clauses, 1 percent of the 1980s clauses, and 2 percent of the 1990s clauses contain language selecting non-contract law.

A screenshot of a cell phone Description automatically generated

Given the relative speed with which the drafting community responded to the late-1970s decisions assigning a narrow ambit to the words “interpret” and “construe,” the lack of any sort of widespread response to the mid-1970s decisions concerning noncontractual law is surprising. The issue was not one of capacity—the courts were very clear that the parties had the power to prospectively select the tort and statutory law of a given jurisdiction.[162] Most companies simply chose not to revise their clauses to take advantage of this opportunity.[163] This inaction is all the more surprising in light of a clause from 1918 that I uncovered during my research. This clause was part of a standard employment contract for circus performers who traveled around the country. It provided that

it is mutually agreed between the parties hereto that the place of the contract . . . is the District of Columbia, [and in said District of Columbia] or according to the laws thereof, if construed or litigated elsewhere, shall all matters whether sounding [in contract] or in tort, relating to its validity, construction, and interpretation be determined to the same extent as if its execution, performance, or cause of action thereon or growing out of the same, actually took place or arose in said District of Columbia.[164]

Stumbling across this particular clause in a contract from 1918 was a bit like opening an ancient Egyptian tomb for the first time and discovering a functioning helicopter inside. The lawyers drafting contracts for this circus company knew enough to prospectively select the applicable tort law decades before other companies began dabbling with the practice in the 1960s and 1970s. The notion that a contract drafter could preemptively select the tort law of a given jurisdiction, in summary, should not have come as a surprise to attorneys in the closing decades of the twentieth century. Nevertheless, the proportion of contracts containing choice-of-law clauses that selected the noncontractual law of the chosen jurisdiction was negligible.

Clauses that Exclude the Conflicts Rules of the Chosen Jurisdiction

If attorneys in the United States were curiously apathetic to the possibilities presented by broad choice-of-law clauses, they were inordinately alarmed by the risks presented by clauses that did not expressly exclude the conflict-of-laws rules of the chosen jurisdiction. To understand these risks—and why they are unlikely to come to pass in practice—it is necessary to take a brief detour into conflict-of-laws doctrine.

Conflicts scholars frequently distinguish between the internal law of a given state, excluding its conflict-of-laws rules, and the whole law of a given state, including its conflict-of-laws rules.[165] This distinction has the potential to generate ambiguity as to what “laws” the parties are referencing in their choice-of-law clauses. Did they mean to select the chosen state’s internal law? Or did they mean to select its whole law? Since the purpose of the choice-of-law clause is to reduce legal uncertainty, it stands to reason that most parties intend to select the internal law of the chosen state. Nevertheless, many choice-of-law clauses do not clearly state which type of “laws” they are choosing.

In the 1940s and 1950s, U.S. courts reached different conclusions as to how this textual ambiguity should be resolved. In 1947, the Sixth Circuit concluded that a choice-of-law clause selecting the “laws” of Pennsylvania referred to the whole law of that state. After reviewing the relevant Pennsylvania conflicts rules, the court concluded that—per the choice-of-law clause—the contract was governed by the law of Alabama.[166] Needless to say, this outcome was probably not the one intended by the parties when they chose Pennsylvania law to govern their agreement.[167] In 1955, by contrast, the Second Circuit was called upon to determine the meaning of the word “laws” in a choice-of-law clause selecting the laws of England. The court concluded that the word “laws” should be read to select the internal law of England because the entire purpose of the clause was to ensure a uniform result regardless of where the litigation occurred.[168] If the clause were interpreted to select the whole law of England, the court reasoned, then this objective could not be obtained because it could potentially lead to the application of the laws of still another nation.

In the wake of these conflicting decisions about the core meaning of an increasingly popular contractual provision, contract drafters should have immediately set about revising their choice-of-law clauses to make clear their intent. If these revisions occurred, they are absent from the historical record. There is not a single clause from the 1950s or the 1960s that addresses the internal law‌/whole law distinction. The first contract that I found selecting the internal law of a state dates to 1970.

While there is no way to know precisely why such clauses first began appearing in the early 1970s, the most likely explanation is that they were responding to the Second Restatement, which was finalized in 1969 and published in 1971. That work made a point to emphasize that the word “law” in choice-of-law clauses should typically be read to refer to the internal law of the chosen state rather than its whole law.[169] In the comments to section 187, for example, the drafters observed that:

When they choose the state which is to furnish the law governing the validity of their contract, the parties almost certainly have the [internal law] rather than the [whole law] of that state in mind. To apply the [whole law] of the chosen state would introduce the uncertainties of choice of law into the proceedings and would serve to defeat the basic objectives, namely those of certainty and predictability, which the choice-of-law provision was designed to achieve.[170]

Even though the position taken by the Second Restatement accorded with the preferences of virtually all parties, and even though this position was followed by virtually every U.S. court to consider the issue, there was still the lingering possibility that a judge might misinterpret the intended meaning of the word “laws” in a choice-of-law clause.[171] To guard against this remote possibility, some contract drafters began to redraft their clauses.[172] In some cases, they selected the “internal” laws of the chosen jurisdiction. In others, they stated that the court was to apply the law of the chosen jurisdiction “without regard to principles of conflict of laws.”[173] In still others, they directed the courts to apply the law that would be “applicable to contracts made and to be performed” in the chosen jurisdiction.[174] Although the language used in each case was different, the impulse was the same—to ensure that the courts did not apply the conflicts rules of the chosen jurisdiction were the contract ever to wind up in litigation.

Although the practice of drafting clauses to exclude the conflicts rules of the chosen jurisdiction began around 1970, such clauses were rarely found in contracts executed over the course of the following decade. Only 1 percent of the clauses from the 1970s contain language excluding the conflicts rules of the chosen jurisdiction.[175] In the 1980s, however, 8 percent of the clauses excluded these rules. This number jumped again in the 1990s, where the issue was addressed in 18 percent of clauses.[176]

A screenshot of a cell phone Description automatically generated

It is something of a mystery why this language caught on in the 1980s. Contract drafters had, after all, failed to respond to a number of other events in prior decades that should, in theory, have triggered a change in drafting practice. Whatever the cause, the data show that, beginning around 1980, an increasing proportion of choice-of-law clauses contained language excluding the conflicts law of the chosen jurisdiction.

Procedural Law

As a general rule, courts have held that choice-of-law clauses operate to select the substantive law of the chosen jurisdiction but not its procedural law. In other words, a choice-of-law clause selecting New York law is deemed to select that state’s (substantive) contract law but not its (procedural) law of evidence. On occasion, however, a choice-of-law clause purports to select the procedural law of the chosen jurisdiction. Some clauses provide, for example, that the agreement shall be “enforced” in accordance with the chosen jurisdiction’s law.[177] Other clauses provide that the “remedies” available to the parties shall be determined by the chosen jurisdiction’s law. Still other clauses state outright that the courts should apply the “procedural law” of the jurisdiction named in the clause.[178] The figure below highlights the relative frequency of clauses containing such language.[179]

A screenshot of a map Description automatically generated

 

Clauses that select the “procedural law” of a given jurisdiction, or that state that the “remedies” available to the parties shall be determined in accordance with that law, are exceedingly rare. By comparison, there are a fair number of clauses that specify that the agreement shall be “enforced” in accordance with the laws of the chosen jurisdiction; the proportion of such clauses has ranged between 10 percent and 20 percent over the past several decades. Such provisions operate primarily to select certain aspects of the chosen jurisdiction’s law—like statutes of limitations—that straddle the line between substance and procedure and to direct the courts to apply these aspects regardless of how they are characterized.[180]

Other Language

In reviewing thousands of choice-of-law clauses, unusual or distinctive provisions inevitably appear. Although such clauses do not warrant extensive analysis, this Section briefly discusses several additional words and phrases that occasionally appear in choice-of-law clauses. Although these unusual or distinctive clauses present interesting conceptual issues, they rarely appeared in contracts executed between 1880 and 2000.

First, there were almost no clauses that expressly incorporated the law of a given state into the contract by reference.[181] While courts have long recognized that parties have the ability to incorporate such law into their agreements, the choice-of-law clauses in the data set rarely did so.[182] Second, there were virtually no clauses that referenced a specific state or federal statute.[183] In the overwhelming majority of cases, the parties simply selected the “law” or “laws” of the chosen jurisdiction without referencing individual statutes by name.[184] Third, the number of clauses that selected the law of a given state at a particular point in time was negligible. A few clauses specified that they were selecting the law as it existed at the time the contract was made.[185] A few other clauses specified that they were selecting the law as it existed at the time of litigation.[186] The overwhelming majority of clauses, however, did not address the issue at all. Fourth, there were very few “floating” choice-of-law clauses in the data set.[187] A floating choice-of-law clause stipulates that the contract shall be governed by the law of State A if suit is brought in State A but that it shall be governed by the law of State B if suit is brought in State B.[188] Such clauses constituted a tiny percentage of the clauses in the data set. Fifth, and finally, there were virtually no clauses that expressly referenced the chosen jurisdiction’s statutes of limitations.[189]

  1. Implications for Contract Scholarship

Legal scholars have long recognized that contract language can be stubbornly resistant to change. Indeed, there is a thriving academic literature that seeks to explain why some contract language is sticky while other language is not. This Part seeks to integrate the insights from the foregoing discussion of choice-of-law clauses into this literature. It first discusses the pitfalls of trying to use the rational-actor model to explain changes to the choice-of-law clause over the years. It then shows how the best-known general theory of contractual change fails to provide a convincing account for the process of evolution and change in the choice-of-law clause context.

  1. The Myth of the Rational Contract Drafter

The choice-of-law clause is a standard piece of boilerplate. As such, it rarely attracts much in the way of attention from the contracting parties.[190] When a contract is being negotiated by two individuals with unequal bargaining power, moreover, the choice-of-law clause is typically offered as a take-it-or-leave-it proposition. The stronger party—almost always the contract drafter—will demand that the agreement be governed by the law of its home jurisdiction. The weaker party must accede to this demand if it wants to complete the transaction. In such cases, a rational drafter has every incentive to draft a one-sided choice-of-law clause that favors its interests. In addition to selecting the law of its home jurisdiction, for example, the rational drafter would be well-advised to prepare a broad choice-of-law clause—one that selects the tort and statutory law of that jurisdiction—to further reinforce the advantages that flow from litigating with reference to a familiar body of law. If the other side is going to agree to the deal anyway, so the argument goes, a drafter would be foolish not to include a broad clause in the agreement.[191]

Past studies have discussed why it is rational for the weaker party to sign a contract that contains these one-sided provisions. Briefly, these studies have found that many contract terms are non-salient to the weaker party.[192] A consumer purchasing a new car, for example, is far more likely to focus on the price of the vehicle and on the warranties offered by the dealer than on the choice-of-law clause in the bill of sale. While this state of affairs may produce inefficient contract terms in the aggregate, it is arguably rational from the perspective of the weaker party.[193] The overwhelming majority of purchases, after all, do not result in a lawsuit or even the threat of a lawsuit.

While the existing literature provides a convincing explanation for why it is generally rational for weaker parties to accept one-sided contract provisions, that literature has surprisingly little to say about why stronger parties sometimes fail to draft provisions that maximize their advantages. If the weaker party is going to enter into the transaction regardless of whether the contract contains a choice-of-law clause—and regardless of whether that clause is narrow or broad—then it is arguably irrational for the contract drafter not to include a broad choice-of-law clause selecting the noncontractual law of its home jurisdiction. This is especially true when the stronger party is utilizing the same template for all transactions of a given type. Contract drafters do not, however, always behave rationally when drafting choice-of-law clauses.

Consider the question of whether to write a choice-of-law clause into a contract in the first place. In the 1880s, a great many insurance companies selling life insurance policies across state lines wrote choice-of-law clauses selecting the law of the insurer’s home jurisdiction into their agreements. This was an eminently rational decision. These clauses were non-salient to prospective policyholders but could provide the insurer with important tactical advantages should the contract ever be litigated. Over the next eighty years, however, most other contract drafters chose not to write these clauses into their agreements. It was not until the early 1960s that these clauses began their slow march into the mainstream. This pattern of practice is very difficult to explain.[194] Choice-of-law clauses simultaneously eliminate the need to perform a conflicts analysis in the event of a dispute and offer manufacturers and other corporate entities the opportunity to lock in the contract law of their home jurisdiction.[195] And yet these provisions were rarely used outside of a few industries in the first half of the twentieth century.

Precisely the same critique may be applied to these corporate entities’ collective failure to write broad choice-of-law clauses selecting noncontractual law into their standard form agreements. As discussed above, there are examples of clauses that select the tort law of a given jurisdiction as early as 1918. In the closing decades of the twentieth century, however, this broad language was virtually never found in choice-of-law clauses. This pattern of practice is exceptionally difficult to explain. It is one thing to add an entirely new provision to an agreement. It is quite another to add a short phrase—“claims related to the agreement”—to an already-existing contract clause when the addition would clearly advance the interests of the contract drafter. And yet this phrase was rarely added.

This omission becomes even more mystifying in light of the fact that many of these same companies did revise their choice-of-law clauses to exclude the conflict-of-laws rules of the chosen jurisdiction. As discussed above, the risk that a modern court would ever interpret the word “laws” in a choice-of-law clause to select the whole law of the chosen jurisdiction, including its conflicts rules, is miniscule. Such a decision would be contrary to mountains of precedent as well as the position laid down by the Second Restatement. And yet the number of clauses that specifically exclude the conflicts rules of the chosen jurisdiction rose significantly in the 1980s and 1990s. While there is certainly no harm in drafting clauses in this manner, it is not entirely clear why this particular innovation caught on while other potentially more useful innovations (like broad clauses selecting noncontractual law) did not. All of this, again, highlights the challenges of applying a rational-actor model to explain the process of contract drafting.

It is tempting to think of contract drafting as a modular process akin to an automobile assembly line, in which language is added or removed in a mechanical and systematic way. While it is understandable why this model appeals to people—it seems like this is how things should work—it does a poor job of explaining the changes to the choice-of-law clause described above. Contract drafters do not always behave rationally when it comes to these clauses. In some cases, they incorporate new language that is arguably unnecessary. In other cases, they omit language that would arguably serve to better protect their interests.[196] The foregoing history, in short, provides a useful corrective to the view that stronger contracting parties inevitably leverage their bargaining power to force weaker parties to accept provisions that favor the drafter. Whether out of kindness or ignorance, contract drafters do not always act in this way.[197]

  1. New Models of Contractual Innovation

In recent years, a number of scholars have sought to develop a theoretical model that seeks to explain how and why contractual innovations occur. The most influential model is one put forth by Stephen Choi, Mitu Gulati, and Eric Posner to explain changes to a specific provision—the collective action clause (CAC)—in sovereign debt agreements.[198] This model is generally known as the “shock model.”

The shock model posits that the process of contractual innovation proceeds in three stages. In stage one, a particular standard form dominates the market. Stage one comes to an end when an external “shock” disrupts the status quo.[199] While the precise nature of this shock is highly variable—it can be legal, political, or economic—the shock serves as the catalyst for the process of change and kicks off stage two.[200] In stage two, marginal law firms begin to experiment with contractual innovations that depart from the dominant standard form. The effect of these actions is to reduce the dominance of the existing standard form. In stage three, elite law firms and other market leaders begin to promulgate their own innovations with the goal of establishing a new standard.[201] These innovations continue until practice gradually coalesces around a single proposal. This innovation then becomes the new standard form and the cycle begins again.

In the highly centralized world of sovereign debt agreements and collective action clauses, this model works. In the highly decentralized world of general contracts and choice-of-law clauses, it works less well.[202] Unlike collective action clauses, choice-of-law clauses appear in many different types of contracts. Unlike collective action clauses, choice-of-law clauses are frequently drafted by lawyers who do not work at large firms.[203] And unlike collective action clauses, choice-of-law clauses are often prepared by lawyers who do not work in New York City.[204] All of these differences call into question the explanatory force of the shock model as applied to contractual innovations outside of the context of sovereign debt agreements. In the clubby world of sovereign debt, it may be possible for a relatively small contracting community to coalesce around a single proposal and to quickly incorporate that language into agreements of that type going forward. In the decentralized world of non-sovereign-debt contracts, by comparison, it is difficult for contract drafters preparing different types of agreements for clients in different industries in different parts of the country to coalesce around much of anything.

To the extent that these disparate actors rally around a common standard, moreover, the process of change is likely to be slower than the one envisioned by the shock model. In the years since that model was first proposed, it has been invoked by a number of other contracts scholars to suggest that, as a general matter, the process of contractual change is rapid and explosive.[205] Matthew Jennejohn, for example, has argued that “boilerplate evolves according to a punctuated equilibrium model of institutional change, by which concentrated and often dramatic adjustments follow long stretches of stasis.”[206] And David Hoffman has invoked the model to argue that “[c]ontract terms do not evolve linearly and progressively in rational counterpoint to slow changes in doctrine. They change contingently, explosively, and at moments punctuated by shocks.”[207]

While the process of contractual change as it relates to collective action clauses may well be quick and dramatic, contractual change proceeds differently in the context of choice-of-law clauses. Between 1930 and 1970, for example, the percentage of such clauses that contained some variation of the word “govern” held relatively constant at around 40 percent. In the 1970s, the percentage rose to 55 percent. In the 1980s, it rose to 68 percent. In the 1990s, it rose again to 73 percent. The process of change was not “explosive” or punctuated by “shocks.” Instead, it was slow, steady, and incremental.[208] The increase in the proportion of choice-of-law clauses that excluded the conflicts rules of the chosen jurisdiction—which went from 1 percent in the 1970s to 8 percent in the 1980s to 18 percent in the 1990s—was similarly incremental. In some cases, it would seem, the process of contractual innovation is more akin to the tide coming in than to a tsunami.

None of this is to argue that the shock model should be tossed to the curb. Nor is it to suggest that contract language never evolves explosively or at moments punctuated by shocks. It is merely to point out that the process of contractual change can vary depending upon a range of variables—including, but not limited to, the nature of the contract provision at issue, the type of agreement, the identity of the drafter, the geographic location where the contract is prepared, and the centralized or decentralized nature of the drafting community—and that a great deal depends upon context. It is important to resist, in other words, the temptation to view the shock model as a general model of contractual innovation. The shock model provides a compelling explanation for changes to collective action clauses in sovereign debt agreements. It does not, however, do a good job at explaining changes to choice-of-law clauses in other types of contracts over time. This insight, in turn, suggests the need for new models of contractual innovation beyond the shock model.

Conclusion

There are many histories of the doctrine of party autonomy. To date, however, scholars have largely ignored the history of the choice-of-law clause, without which there would be no need for this doctrine. This Article has sought to tell, for the first time, the history of this increasingly common piece of contract boilerplate. It has tracked the prevalence of these clauses over time and has chronicled the interplay between drafters, courts, scholars, and legislatures concerning their enforceability. It has sought to determine how the language in these clauses has evolved and changed over many years. And it has drawn upon the foregoing history to challenge the myth of the rational contract drafter and to call for more differentiated models of contractual innovation. Whether it has achieved each of these goals is for the reader to determine. At a minimum, however, this Article offers answers to a host of basic questions relating to the choice-of-law clause for which the existing literature offers no ready answers.

 

Appendix: Published Case Data Collection

In assembling the data set of 3,104 choice-of-law clauses discussed in Part III of this Article, the following data collection methods were used.

Each research assistant was given a list of search terms keyed to words and phrases often found in choice-of-law clauses and a specific time frame in which to search.[209] They then searched for the terms in the “Cases” database maintained by Lexis Advance for the assigned time frame. When the research assistant found a case containing a choice-of-law clause, he or she entered the name of the case, the year of the contract containing the choice-of-law clause, the type of contract, and the full text of the choice-of-law clause into a spreadsheet. I subsequently reviewed each spreadsheet entry to confirm that the language in question was, in fact, a choice-of-law clause. I also conducted random spot checks to verify the contract year and the type of agreement.

In the overwhelming majority of instances, the clauses identified as part of this search did not play a meaningful role in the court’s analysis. The court simply quoted the text from the clause as part of its background discussion and then moved on to discuss matters unrelated to choice of law.

When the published case did not state the year in which the contract containing the choice-of-law clause was concluded, this particular cell was left blank. After all of the data collection for a particular decade was complete, I calculated the average gap between the year of the case and the year of the contract for those contracts where the date of execution was known. I then assigned the contracts for which the year was unknown an execution date that matched the average for all of the known agreements. With respect to a choice-of-law clause in a contract with an unknown date from a case decided in 1959, for example, the contract was assigned a year of 1951 because the average difference between the date of the case and the date of the known contracts for cases in the 1950s was eight years. With respect to a choice-of-law clause contained in a contract with an unknown date from a case decided in 1934, by contrast, the contract was assigned a year of 1928 because the average difference between the date of the case and the date of the known contracts for cases in the 1930s was six years.

Once the contracts had all been assigned to a decade, I reviewed the language in each of the clauses so that it could be properly coded. In some instances, this task was easy. With respect to the issue of interpretation, for example, I simply counted how many clauses contained variations on the word “interpret.” In other instances, this task was more complicated. With respect to the issue of internal law, for instance, there are many different ways of signaling an intent to select a given state’s internal law. For these clauses, I had to review each clause individually to see if it contained the phrases “internal law,” “without regard to principles of conflict of laws,” “applicable to contracts made and to be performed,” or some other language to similar effect. After all of the clauses had been coded, I then calculated the percentage of clauses for a given decade that contained language responsive to the issue under discussion.

Although I collected data for clauses going back to the nineteenth century, most of the analysis in Part III focuses on clauses from the 1940s onward. There are several reasons for this. First, the clauses dating to the 1880s, 1890s, 1900s, and 1910s were overwhelmingly from a single type of contract—life insurance agreements—and were generally drafted in the same manner. Any general discussion of the clauses from these decades, therefore, would give a misleading impression of broader contract practice because the sample was so dominated by life insurance agreements. Second, the total number of clauses from these early decades was generally small. There were only thirty-one clauses from the 1910s, forty-three clauses from the 1920s, and fifty-three clauses from the 1930s. From that point forward, the number of clauses in the sample increased substantially. There were seventy-nine clauses from the 1940s, 123 clauses from the 1950s, and 236 clauses from the 1960s. Accordingly, the presentation of the data focuses principally on those decades for which there was a larger—and presumably more reliable—body of contracts.

A critic might argue that the sample of clauses here examined is biased because all of them come from published cases. If published cases are not representative of all cases—and they are not—then it may well be that choice-of-law clauses quoted in published cases are not representative of all choice-of-law clauses.[210] This critique is entirely fair. If the goal is to study the changes to the language in choice-of-law clauses over the course of the twentieth century, however, this particular method of data collection is arguably the best of all the bad options. In a perfect world, I would spend a year traveling from company archive to company archive in search of original paper contracts from companies across a wide range of industries in search of choice-of-law clauses. Given the enormous number of contracts negotiated over the past 150 years, however, it is not clear that the sample of clauses gathered through such an endeavor would be any more representative than the ones collected from a review of published cases. Such an inquiry would inevitably omit contracts from companies that failed. It would also omit contracts from companies that do not maintain archives open to researchers.

At the end of the day, the simple truth is that the vast majority of all historical contracts are unavailable to researchers. This means that any attempt to reconstruct past contracting practice must inevitably rely on a sample that may not be representative. This discussion calls to mind the old story of the man looking for his car keys under the streetlamp because it is the only place where he can see to look. While searching through published cases for choice-of-law clauses may be less than ideal, it is the one of the very few places where we can see to look.

  1. * Reef C. Ivey II Distinguished Professor of Law, University of North Carolina at Chapel Hill. Thanks to Ronald Brand, Pam Bookman, Christian Burset, Lauren Coyle, Bill Dodge, Maggie Gardner, Mitu Gulati, Carissa Hessick, Dave Hoffman, Alex Mills, Tanya Monestier, Leigh Osofsky, Steve Sachs, Aaron Simowitz, Kathleen Thomas, Mark Weidemaier, Glenn West, Chris Whytock, and the faculty of the Sandra Day O’Connor School of Law at Arizona State University for comments on an earlier draft of this Article. Thanks to Will Bowers, Kelsey Cullinan, Katie Dixon, Ryan Dovel, Rashaad Hamilton, Richard Hanson, Trevor Kennedy, Ker Medero, Foram Majmudar, M-K McKinney, Marissa Mugan, Jordan Proffitt, Samantha Reeves, Alexandra Russell, Stephen Spivey, Luke Thomas, Andrew Wisniewsky, Edward Woodall, and Christine Xiao for their excellent research assistance. Thanks to Harriet Wu for her help with the figures and tables.
  2. . See Restatement (Second) of Conflict of Laws § 188 (Am. Law Inst. 1971).
  3. . See, e.g., Janet V. Hallahan, The Case of the Missing Decision: When Will Pennsylvania Solve the Mystery of Its “Flexible” Choice-of-Law Analysis?, 69 Temp. L. Rev. 655, 694 (1996).
  4. . See, e.g., Peter Hay, The Use and Determination of Foreign Law in Civil Litigation in the United States, 62 Am. J. Comp. L. 213, 217 (2014); Christopher A. Whytock, Myth of Mess? International Choice of Law in Action, 84 N.Y.U. L. Rev. 719, 732–33 (2009).
  5. . See Mastrobuono v. Shearson Lehman Hutton, Inc., 514 U.S. 52, 59 (1995) (observing that a “choice of law provision, when viewed in isolation, may reasonably be read as merely a substitute for the conflicts-of-laws analysis that otherwise would determine what law to apply to disputes arising out of the contractual relationship”). A recent study found support for the concept of party autonomy in conflict of laws in legal systems all around the world. See Symeon Symeonides, The Scope and Limits of Party Autonomy in International Contracts: A Comparative Analysis, in The Continuing Relevance of Private International Law and New Challenges (Franco Ferrari & Diego P. Fernandez Arroyo eds., 2019).
  6. . Julian Nyarko, We’ll See You in . . . Court! The Lack of Arbitration Clauses in International Commercial Contracts, 58 Int’l Rev. L. & Econ. 6, 11 (2018).
  7. . Interestingly, the history of the forum selection clause has garnered more academic attention than the history of the choice-of-law clause notwithstanding the fact that the latter was (and is) more common than the former. See David Marcus, The Perils of Contract Procedure: A Revised History of Forum Selection Clauses in the Federal Courts, 82 Tul. L. Rev. 973, 996–1014 (2008) (discussing the history of the forum selection clause).
  8. . See, e.g., Alex Mills, Party Autonomy in Private International Law 44–64 (2018); Symeon Symeonides, Codifying Choice of Law Around the World: An International Comparative Analysis 110–15 (2014); Matthias Lehman, Liberating the Individual from Battles Between States: Justifying Party Autonomy in Conflict of Laws, 41 Vand. J. Transnat’l L. 381, 385–90 (2008); Gisela Ruhl, Party Autonomy in the Private International Law of Contracts: Transnational Convergence and Economic Efficiency, in Conflict of Laws in a Globalized World 4–8 (Eckart Gottschalk ed., 2007).
  9. . Similarly, there is no need for any doctrine of party autonomy with respect to the chosen forum absent a forum selection clause or an arbitration clause.
  10. . As an empirical matter, it is a near-impossible task to provide a comprehensive assessment of contract drafting practice as it existed more than a century ago. Each of the sources listed in this paragraph, therefore, represents a second-best attempt to determine what such practice looked like. Collectively, they constitute the most comprehensive data set possible that can help us make inferences about party behavior, notwithstanding the limitations inherent in their use.
  11. . Joseph Story, Commentaries on the Conflict of Laws § 233 (1834).
  12. . Id.
  13. . Id.
  14. . See id. § 278.
  15. . Robinson v. Bland, Eng. Rep. 717 (K.B. 1760) (emphasis added).
  16. . Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 48 (1825) (Marshall, C.J.) (emphasis added); see also Morris J. Levin, Party Autonomy: Choice-of-Law Clauses in Commercial Contracts, 46 Geo. L.J. 260, 269 (1957).
  17. . See infra note 19 and accompanying text (discussing the first express choice-of-law clauses); see also Carnegie v. Morrison, 43 Mass. 381, 401 (1841) (“There is no reference, tacit or express, in this instrument, to the laws of England, which can raise a presumption, that the parties looked to them as furnishing the rule of law, which should govern this contract.” (emphasis added)).
  18. . There are a small number of cases in which the court relied upon writings in non-usury contracts to ascertain the intent of the parties. See, e.g., Strawbridge v. Robinson, 10 Ill. 470, 472–73 (1849) (concluding that the parties had expressly selected the law of Illinois by writing “East Fork the 4th Feb’y 1845” on the face of their agreement).
  19. . Butters v. Olds, 11 Iowa 1, 2–3 (1860); see also Scott v. Perlee, 39 Ohio St. 63, 66–67 (Ohio 1883) (citations omitted) (“[I]t is undoubtedly the law of this state, and indeed it is now well established almost universally, that where a contract is entered into in one state, to be performed in another, between citizens of each, and the rate of interest is different in the two, the parties may, in good faith, stipulate for the rate of either, and thus expressly determine with reference to the law of which place that part of the contract shall be decided.”); Peck v. Mayo, Follett & Co., 14 Vt. 33, 38 (1842) (“If a contract be entered into in one place to be performed in another, and the rate of interest differ in the two countries, the parties may stipulate for the rate of interest of either country, and thus, by their own express contract, determine with reference to the law of which country that incident of the contract shall be decided.”).
  20. . Kirtland v. Hotchkiss, 42 Conn. 426, 444 (1875). This is not to suggest that the express choice-of-law clause sprang—like Athena—fully formed from the head of Zeus in 1869. There are clauses in English contracts from the 1760s directing arbitrators to apply the “Rules and Customs” of a particular place. See Christian Burset, Arbitrating the England Problem: Litigation, Private Ordering, and the Rise of the Modern Economy, 36 Ohio St. J. on Disp. Resol. (forthcoming 2020), https://ssrn.com/abstract=3561136 [https://perma.cc/B8DT-V5HP]. And choice-of-law clauses can be found in foreign maritime contracts dating to 1800 that select the law of a particular place. See Aertsen v. The Aurora, 1 F. Cas. 206, 207 (D.S.C. 1800) (Case No. 95) (referencing document selecting the “marine law of Hamburgh” to regulate “the conduct of officers and seamen aboard vessels belonging to that place”). It is merely to state that the express choice-of-law clause in its modern incarnation dates to roughly 1869.
  21. . Kirtland, 42 Conn. at 444; see also Warner v. Warner, 235 Ill. 448, 456 (1908) (referencing a choice-of-law clause in an 1874 prenuptial agreement).
  22. . Union Cent. L. Ins. Co. v. Pollard, 94 Va. 146, 151 (1896).
  23. . New York Life Ins. Co. v. Cravens, 178 U.S. 389, 395 (1900).
  24. . Mutual Ben. Life. Ins. Co. v. Robison, 54 F. 580, 582 (N.D. Iowa 1893).
  25. . The Oranmore, 24 F. 922, 923 (D. Md. 1885). The judge in this case specifically noted that the express choice-of-law clause was a provision “which I have not met with before.” Id.
  26. . The Kensington, 183 U.S. 263, 269 (1902).
  27. . Id.
  28. . New Eng. Mortg. Sec. Co. v. Vader, 28 F. 265, 275 (C.C.D. Or. 1886); see also Farrior v. New Eng. Mortg. Sec. Co., 88 Ala. 275, 277 (1889) (stating that “the notes herein described, and this mortgage, shall be governed and construed by and under the laws of the State of Alabama, where the same is made”).
  29. . Padley v. Neill, 134 Mo. 364, 369–70 (1896).
  30. . Smith v. Parsons, 55 Minn. 520, 526 (1893).
  31. . In 1882, the Supreme Court reiterated its support for the general principle of party autonomy in a case that did not contain an express choice-of-law clause. See Pritchard v. Norton, 106 U.S. 124, 137 (1882) (“It is always to be remembered that in obligations it is the will of the contracting parties, and not the law, which fixes the place of fulfilment—whether that place be fixed by express words or by tacit implication—as the place to the jurisdiction of which the contracting parties elected to submit themselves.”).
  32. . See Lanier v. Union Mortg. Banking & Tr. Co., 40 S.W. 466, 469 (Ark. 1897); Wilson v. Lewiston Mill Co., 150 N.Y. 314, 323 (1896); Union Cent. Life Ins. v. Pollard, 26 S.E. 421, 422 (Va. 1896); Smith v. Parsons, 57 N.W. 311, 313 (Minn. 1893); New Eng. Mortg. Sec. Co. v. McLaughlin, 13 S.E. 81, 83 (Ga. 1891).
  33. . Sometimes the problems were more prosaic. In 1900, for example, the Supreme Court observed that “a contract by an insurance company of New York executed elsewhere may by its terms incorporate the law of New York, and make its provisions controlling upon both the insured and the insurer.” Mutual Life Ins. v. Cohen, 179 U.S. 262, 267 (1900). On the facts of the case before it, however, the Court held that because the choice-of-law clause had been written into the application for a life insurance policy rather than the policy itself, the parties had not, in fact, chosen the law of New York to govern their agreement. Id. at 270.
  34. . Fletcher v. N.Y. Life Ins. Co., 13 F. 526, 528 (1882) (emphasis added); see also Francis Wharton, A Treatise on the Conflict of Laws 1194 (1905) (“[I]f a borrower residing in one state and a lender in another enter into a contract in the latter state, making it payable in the former, and fix a rate of interest greater than allowed by the law of either state, but expressly stipulate that the contract shall be governed by the law of a third state where such rate is legal, it is apparent, at least in the absence of additional circumstances showing that some of the elements of the contract or significant circumstances of the transaction have their situs in that state, that the parties acted in bad faith and for the purpose of evading the usury law of one or both of the other states.”).
  35. . The Energia, 56 F. 124, 127 (S.D.N.Y. 1893) (citations omitted).
  36. . Lindenberger Cold Storage & Canning Co. v. J. Lindenberger, Inc., 235 F. 542, 547 (W.D. Wash. 1916) (stipulating that the agreement was “to be construed and take effect as a contract made in England and in accordance with the law of England”).
  37. . Weil v. Chi. Pneumatic Tool Co., 212 S.W. 313, 314 (Ark. 1919) (agreement stipulating that “[it] shall be interpreted and construed according to the Laws of the State of Illinois”).
  38. . Chalmers & Williams v. Surprise, 123 N.E. 841, 842 (Ind. App. 1919) (stating that “the contract shall be deemed consummated at Chicago, Illinois”).
  39. . Carl Hagenback & Great Wallace Show Co. v. Randall, 126 N.E. 501, 502 (Ind. 1920).
  40. . Id.; see also Pollak v. Danbury Mfg. Co., 131 A. 426, 428 (Conn. 1925) (discussing choice-of-law clause written into the employment contract of a general manager at a factory); Laure v. Singer, 100 N.J.L. 98, 100 (N.J. 1924) (discussing choice-of-law clause written into the employment contract of a theatrical performer).
  41. . See Shafer v. Metro-Goldwyn-Mayer Distrib. Corp., 172 N.E. 689, 691 (Ohio 1929); see also Erie, Bernhardt, and Section 2 of the United States Arbitration Act: A Farrago of Rights, Remedies, and a Right to a Remedy, 69 Yale L.J. 847, 854 n.41 (1960) (discussing history of the New York arbitration law that provided the basis for the Federal Arbitration Act).
  42. . See, e.g., Binderup v. Pathe Exch., Inc., 263 U.S. 291, 302 (1923).
  43. . See Warner v. Warner, 85 N.E. 630, 633–34 (Ill. 1908) (referencing a choice-of-law clause in an 1874 prenuptial agreement).
  44. . Swan v. Swan’s Ex’r, 117 S.E. 858, 859 (Va. 1923).
  45. . Id.
  46. . Boggs v. Boggs, 114 A. 474, 475 (Md. 1921).
  47. . Id.
  48. . Id.
  49. . In the contract form books from this era, for example, less than 2 percent of the agreements contain choice-of-law clauses. See infra Section I.F (discussing form books from the 1910s).
  50. . Although the first life insurance contract to contain a choice-of-law clause dates to 1880, and although most life insurance contracts contained these clauses in 1902, many life insurance contracts issued in the years after 1905 lacked choice-of-law clauses. Compare Allen J. Flitcraft, Life Insurance Manual (15th ed. 1902) (collecting forty-eight life insurance agreements and applications, of which 66 percent contained choice-of-law clauses) with Allen J. Flitcraft, Life Insurance Manual (23rd ed. 1910) (collecting fifty-nine life insurance agreements or applications, of which 7 percent contained choice-of-law clauses). This shift in contract practice is attributable to the enactment of state laws between 1906 and 1909 mandating that insurance companies use standard policy forms approved by the state. See R. Carlyle Buley, I The American Life Convention, 1906–1952: A Study in the History of Life Insurance 274, 280, 310 (1953). These standard policy forms generally lacked choice-of-law clauses. The state legislation thus helps to explain why some of the best known conflicts decisions rendered by the U.S. Supreme Court in the twentieth century involved insurance contracts that lacked a choice-of-law clause notwithstanding the fact that these clauses could be commonly found in policies issued earlier in the century. See, e.g., Allstate Ins. Co. v. Hague, 449 U.S. 302 (1981); Clay v. Sun Ins. Office, 377 U.S. 179 (1964).
  51. . Kirtland v. Hotchkiss, 42 Conn. 426, 444 (1875).
  52. . Gile v. Interstate Motor Car Co., 145 N.W. 732, 733 (N.D. 1914).
  53. . U.S. Mortg. Co. v. Sperry, 138 U.S. 313, 338 (1891).
  54. . Lindenberger Cold Storage & Canning Co. v. J. Lindenberger, Inc., 235 F. 542, 547 (W.D. Wash. 1916).
  55. . Warner v. Warner, 85 N.E. 630, 633–34 (Ill. 1908).
  56. . Binderup v. Pathe Exch. Inc., 263 U.S. 291, 302 (1923).
  57. . Equitable Life Assurance Soc. v. Clements, 140 U.S. 226 (1891).
  58. . Chalmers & Williams v. Surprise, 123 N.E. 841, 842 (Ind. App. 1919).
  59. . The Oranmore, 24 F. 922, 923 (D. Md. 1885).
  60. . Weisberg v. Hunt, 131 N.E. 471, 472 (Mass. 1921).
  61. . The Kensington, 183 U.S. 263, 269 (1902).
  62. . Carl Hagenbeck & Great Wallace Show Co. v. Randall, 126 N.E. 501, 502 (Ind. App. 1920).
  63. . Comm’r v. Bateman, 127 F.2d 266, 268 (1st Cir. 1942).
  64. . State v. Hall, 114 S.E. 250, 251 (W. Va. 1922).
  65. . Thoms v. Thoms, 222 Ill. App. 618, 622 (1921).
  66. . R.C.A. Photophone, Inc. v. Sinnott, 30 P.2d 761, 762 (Or. 1934).
  67. . Samuel Williston, Joseph Henry Beale: A Biographical Sketch, 56 Harv. L. Rev. 685 (1943); see also Symeon Symeonides, The First Conflicts Restatement Through the Eyes of the Old: As Bad as Its Reputation?, 32 S. Ill. U. L.J. 39, 41–45 (2007).
  68. . Joseph H. Beale, What Law Governs the Validity of a Contract, 23 Harv. L. Rev. 1 (1909). Beale argued that “the doctrine itself is one that is quite foreign to common-law notions.” Id. at 7. He argued further that the idea that contracting parties should be allowed to choose a law to govern their contracts “is not a natural notion in a law based like ours on the complete jurisdiction of the territorial sovereign.” Id.
  69. . Joseph H. Beale, II The Conflict of Laws 1095 (1935).
  70. . Id. at 1096.
  71. . Id. at 1105–09 (identifying several U.S. Supreme Court decisions in which the Court resolved a choice-of-law issue without specifically stating that the intent of the parties should be considered).
  72. . Id. at 1085–86.
  73. . Joseph Beale, A Treatise on the Conflict of Laws 107 (1916) (“A right having been created by the appropriate law, the recognition of its existence should follow everywhere.”).
  74. . See Ala. Great S. R.R. v. Carroll, 97 Ala. 126 (1892).
  75. . Restatement (First) of Conflict of Laws § 1 (Am. Law Inst. 1934).
  76. . Beale, supra note 68, at 1079–80.
  77. . Hay et al., Conflict of Laws 21 (5th ed. 2010).
  78. . See Symeonides, supra note 66, at 66–74 (discussing the drafting process within the American Law Institute that led to the production of the First Restatement and explaining how Beale managed to push through a draft that was inconsistent with the case law in many respects).
  79. . The one concession that Beale was willing to make involved contract interpretation. Even he acknowledged that “the parties’ intention governs in this matter of the meaning of the words used.” Beale, supra note 68, at 1201. He believed, however, that interpretation isa wholly different point from that involved in allowing them to choose the law to govern the obligation. Here there is no question of validity; the sole question is that of fixing the meaning of language and there is no more difficulty in giving effect to a Texas use of a term in a Missouri contract than in giving effect to French or German instead of English as a means of expression.Id. These observations were made in his treatise. They were not expressed in the First Restatement.
  80. . Id. at 1174.
  81. . Id.
  82. . Arthur Nussbaum, Conflict Theories of Contracts: Cases Versus Restatement, 51 Yale L.J. 893 (1942).
  83. . Id. at 895. “Private international law” is the phrase that non-U.S. legal scholars use to refer to the body of doctrine that is known as conflict of laws in the United States. See Zicherman v. Korean Air Lines Co., 516 U.S. 217, 227 (1996).
  84. . Nussbaum, supra note 81.
  85. . Id. at 922.
  86. . In 1928, Judge Finch warned Beale that the judges would not blindly follow Beale’s position on this issue even if he wrote it into the First Restatement. See 6 American Law Institute Proceedings 458 (1927–28) (“[Y]‌ou will never be able to hold your courts to that sort of a rule. You can lay it down, but human nature is not so constituted that you can make a court adopt a general rule which will do injustice in a majority of the cases coming with it.”) (quoted in Symeonides, supra note 66, at 113); see also Mills, supra note 7, at 61 (“The rejection of party autonomy, even by such influential figures, did not however halt its rise in practice.”).
  87. . Boseman v. Conn. Gen. Life Ins., 301 U.S. 196, 202 (1937). At about this same time, the U.S. Supreme Court issued several other decisions making clear that the U.S. Constitution did not mandate a reliance on the vested rights theory in choice of law. See James Y. Stern, Choice of Law, the Constitution, and Lochner, 94 Va. L. Rev. 1509, 1525 (2008).
  88. . Hal Roach Studios, Inc. v. Film Classics, Inc., 156 F.2d 596, 598 (2d Cir. 1946) (internal citations and quotation marks omitted). This decision had the effect of overturning two prior Second Circuit decisions authored by Judge Learned Hand criticizing the practice of party autonomy. See E. Gerli & Co. v. Cunard S.S. Co., 48 F.2d 115 (2d Cir. 1931); Louis-Dreyfus v. Paterson S.S., Ltd., 43 F.2d 824 (2d Cir. 1930).
  89. . See also Lauritzen v. Larsen, 345 U.S. 571, 588–89 (1953) (“[T]his contract was explicit that the Danish law and the contract with the Danish union were to control. Except as forbidden by some public policy, the tendency of the law is to apply in contract matters the law which the parties intended to apply. We are aware of no public policy that would prevent the parties to this contract, which contemplates performance in a multitude of territorial jurisdictions and on the high seas, from so settling upon the law of the flag-state as their governing code.” (emphasis added)).
  90. . U.C.C. § 1-105 (Am. Law Inst. 1952).
  91. . This is not the first article to look to form books to answer other questions about historical contract practice. See Lee C. Buchheit & G. Mitu Gulati, Sovereign Bonds and the Collective Will, 51 Emory L.J. 1317, 1325 (2004) (relying on English form books to determine the first use of the collective action clause).
  92. . Adam Freedman, The Party of the First Part: The Curious World of Legalese 25 (2007) (“Almost as soon as Gutenberg’s first Bible rolled off the press (1455), English lawyers were putting together formbooks, that is, collections of sample contracts, pleadings, and other documents that had already passed muster with some court or another. Provided that one copied the form verbatim, no sporting judge could object.”); M.H. Hoeflich, Law Blanks & Form Books: A Chapter in the Early History of Document Production, 11 Green Bag 2d 189, 191 (2008) (“The time-saving aspect of the use of forms was particularly important to American lawyers for much of the nineteenth century because they tended to bill by the transaction, rather than by the hour or increment thereof.”).
  93. . See William E. Foster & Andrew L. Lawson, When to Praise the Machine: The Promise and Perils of Automated Transactional Drafting, 69 S.C. L. Rev. 597, 613–23 (2018) (discussing the changing drafting practices of corporate lawyers over time and the changing role of form books in the computer age).
  94. . Graybill v. Graybill, 14 Pa. D. & C. 382, 384 (C.P. 1930).
  95. . See Nyarko, supra note 5, at 11.
  96. . J.R. Sayler, Sayler’s American Form Book (1913) (250 contracts; 0 clauses); Francis B. Tiffany, Legal and Business Forms (1915) (221 contracts; 2 clauses).
  97. . Clarence F. Birdseye, Encyclopaedia of General Business and Legal Forms (1924) (636 contracts; 11 clauses); Leslie M. O’Connor, Pocket Manual of Standard Legal and Business Forms (1921) (95 contracts; 1 clause); William Herbert Page, The Law of Contracts (1922) (170 contracts; 1 clause).
  98. . Edmund O. Belsheim, I Modern Legal Forms 413–554 (1938); Edmund O. Belsheim, IV Modern Legal Forms 1–438 (1938) (263 total contracts; 6 total clauses).
  99. . Saul Gordon, Gordon’s Modern Annotated Forms of Agreement (1946) (431 contracts; 11 clauses).
  100. . Leonard A. Jones, Legal Forms Annotated: Contractual Business and Conveyancing Forms (1946) (678 contracts; 3 clauses).
  101. . Even if the publication of the First Restatement had generated widespread concern about the enforceability of choice-of-law clauses, there was arguably no real harm to leaving existing clauses in agreements. After all, if a court refused to enforce them, then the parties were no worse off than before.
  102. . I Am. Jur. Legal Forms Annotated (1953) (132 contracts; 1 clause).
  103. . William J. Casey, Forms of Business Agreements with Tax Ideas (1961).
  104. . David W. Carroll, Forms for Commercial Transactions Under the Uniform Commercial Code (1963).
  105. . While it is also possible that the tremendous expansion of commerce after the Second World War led to more contracts between commercial actors in different states and that these dealings with out-of-state parties, in turn, may have led a growing number of parties to write choice-of-law clauses into their agreements, it is impossible to know for sure the extent to which the expansion of commerce during these years contributed to the more extensive use of choice-of-law clauses.
  106. . See Robert J. Nordstrom & Dale B. Ramerman, The Uniform Commercial Code and the Choice of Law, 1969 Duke L.J. 624, 636 n.45 (1969).
  107. . It is unlikely, however, that this increase was driven by a new belief that such clauses were enforceable. As discussed above, the caselaw post-1934 had broken decisively in favor of enforcement such that it would have been unreasonable for contract drafters to consciously omit such clauses from their contracts. It is far more likely that the furor relating to the enactment of the UCC—and the attention given to section 1-105—enhanced the perceived salience of choice-of-law clauses. More lawyers were thinking about choice-of-law clauses because they were made aware of section 1-105. This enhanced salience, in turn, may have prompted attorneys to revise their standard forms to incorporate choice-of-law clauses.
  108. . Restatement (Second) of Conflict of Laws § 187(1) (Am. Law Inst. 1971). A default rule is one that the parties can contract around in their agreement. The parties may, for example, select the law of any state to govern issues relating to contract interpretation because they could just as easily rewrite their agreement to resolve the interpretive issue themselves.
  109. . Id. § 187(2). A mandatory rule is one that the parties cannot contract around in their agreement. The parties generally may not, for example, select the law of just any state to govern issues relating to securities laws because most states require that their securities laws be applied to certain transactions.
  110. . Id.
  111. . Id. cmt. f.
  112. . Id.
  113. . Id. cmt. g; see also supra notes 33–34 and accompanying text.
  114. . In the years since the Second Restatement was published, section 187 has proven to be popular among U.S. courts. Hay et al., supra note 76, at 75 (describing section 187 as “one of the Restatement’s most successful and popular provisions”). Even when a particular jurisdiction has declined to follow the Second Restatement with respect to other issues, its courts will frequently look to section 187 to determine the enforceability of a choice-of-law clause. Id. at 1088 (observing that section 187 “is followed by more American courts than any other provision in the Restatement (Second), including some courts that otherwise follow the traditional theory”) (internal quotation marks and citations omitted).
  115. . The statute provided:if a policyholder had made at least two timely payments of premiums, defaulted on a subsequent premium, and died within sixty days of the default, the beneficiary [could] recover the accrued value of the policy “as if there had been no default in the payment of premium, anything in the policy to the contrary notwithstanding.”Clyde Spillenger, Risk Regulation, Extraterritoriality, and Domicile: The Constitutionalization of American Choice of Law, 1850–1940, 62 UCLA L. Rev. 1240, 1298 (2015) (emphasis added).
  116. . See N.Y. Life Ins. v. Cravens, 178 U.S. 389, 401 (1900). Even today, a few states have laws on the books that require state courts to apply state law to all contracts insuring lives, property, or interests within the state. See, e.g., N.C. Gen. Stat. Ann. § 58-3-1 (West 2019); S.C. Code Ann. § 38-61-10 (West 2019); Va. Code. Ann. § 38.2-313 (West 2019).
  117. . See Unif. Consumer Credit Code § 1.201(8). This Code has been adopted by eleven states.
  118. . See George F. Carpinello, Testing the Limits of Choice of Law Clauses: Franchise Contracts as a Case Study, 74 Marq. L. Rev. 57, 71 n.68 (1990) (listing states that have enacted franchise statutes containing anti-waiver provisions).
  119. . See, e.g., Cal. Lab. Code § 925(a) (West 2019) (“An employer shall not require an employee who primarily resides and works in California, as a condition of employment, to agree to a provision that would . . . [d]eprive the employee of the substantive protection of California law with respect to a controversy arising in California.”); Kan. Stat. Ann. § 50-625(a) (West 2019) (“Except as otherwise provided in this act, a consumer may not waive or agree to forego rights or benefits under this act.”); see also La. Stat. Ann. § 23:921(A)(2) (West 2019) (“The provisions of every employment contract or agreement . . . by which any foreign or domestic employer or any other person or entity includes a . . . choice of law clause in an employee’s contract of employment . . . shall be null and void except where the . . . choice of law clause is expressly, knowingly, and voluntarily agreed to and ratified by the employee after the occurrence of the incident which is the subject of the civil or administrative action.”).
  120. . See generally Hay et al., supra note 76, at 1108–26 (discussing public policy limitations across a range of different types of agreements).
  121. . See Cyra Akila Choudhury, Shari’ah Law as National Security Threat?, 46 Akron L. Rev. 49, 52–65 (2013).
  122. . See, e.g., Ark. Code Ann. § 1-1-103 (2019). In some cases, the legislation distinguished between commercial contracts and noncommercial contracts. See N.C. Gen. Stat. § 1-87.14 (2019) (“A court . . . shall not apply a foreign law in any legal proceeding involving . . . a claim for absolute divorce, divorce from bed and board, child custody, child support, alimony, or equitable distribution if doing so would violate a fundamental constitutional right of one or more natural persons who are parties to the proceeding.”). In other cases, it did not. See Kan. Stat. Ann. § 60-5104 (2019). In a number of instances, the legislation provided that the prohibition on applying foreign law would not apply to artificial entities such as corporations. See, e.g., Ariz. Rev. Stat. Ann. § 12-3102 (2019); La. Stat. Ann. §9:6001(G) (2019); Kan. Stat. Ann. § 60-5108 (2019).
  123. . Ex parte Cont’l Motors, Inc., 270 So. 3d 1148, 1151 (Ala. 2018).
  124. . N.Y. Gen. Oblig. L. § 5-1401(1) (McKinney 1984). The statute stated that it would not apply to any contract for labor or personal services, any contract for personal, family, or household services, or contracts implicating issues covered by certain sections of the Uniform Commercial Code. Id. § 5-1401(2).
  125. . See Theodore Eisenberg & Geoffrey P. Miller, The Market for Contracts, 30 Cardozo L. Rev. 2073, 2091 (2009).
  126. . See, e.g., Cal. Civ. Code § 1646.5 (West 2019); Del. Code Ann. Tit. 6, § 2708 (West 2019); Fla. Stat. Ann. § 685.101 (West 2019); 735 Ill. Comp. Stat. Ann. 105‌/5-5 (West 2019); Ohio Rev. Code Ann. § 2307.39 (West 2019).
  127. . N.C. Gen. Stat. Ann. § 1G-3(a)(1) (West 2019). The statute defines a business contract as “a contract or undertaking, contingent or otherwise, entered into primarily for business or commercial purposes. The term does not include a consumer contract or employment contract.” Id. § 1G-2.
  128. . Id. § 1G-3(a)(2).
  129. . .Sarath Sanga, Choice of Law: An Empirical Analysis, 11 J. Empirical Legal Stud. 894, 90203 (2014).
  130. . Nyarko, supra note 5, at 10–11.
  131. . John F. Coyle & Christopher R. Drahozal, An Empirical Study of Dispute Resolution Clauses in International Supply Contracts, 52 Vand. J. Transnat’l L. 323 (2019); see also Gilles Cuniberti, The International Market for Contracts: The Most Attractive Contract Laws, 34 Nw. J. Int’l L. & Bus. 455, 459, 469 (2014) (“A study of contracts in cases that came before the International Court of Arbitration of the International Chamber of Commerce between 2007 and 2012 found that 83% contained choice-of-law clauses.”).
  132. . Kirtland v. Hotchkiss, 42 Conn. 426, 444 (1875). In order to appreciate how little has changed in some respects since the 1860s, compare this clause to the following clause from 2015: “[This Agreement] shall be deemed to be a contract made under the laws of the State of New York, and for all purposes shall be governed by and construed in accordance with the laws of said State.” Carnival PLC and U.S. Bank Nat’l Ass’n, Indenture (Form 305B2) (Nov. 6, 2015); see also Fairmount Santrol Holdings Inc. and The Bank of N.Y. Mellon Tr. Co., Indenture (Form 305B2) (Dec. 1, 2016) (“This Indenture and the Securities will be deemed to be a contract made under the laws of the State of New York, and for all purposes will be construed in accordance with the laws of said State without giving effect to principles of conflicts of laws of such State.”).
  133. . We also collected 243 clauses that selected the law of a foreign jurisdiction. Since this Article is focused exclusively on U.S. contract practice, and since most choice-of-law clauses selecting a foreign jurisdiction were drafted by foreign parties with an eye to foreign law, these foreign clauses were excluded from the data set.
  134. . Bailey v. Jerome, 129 S.C. 387 (1924) (emphasis added); see also Farmers’ & Breeders’ Mut. Reserve Fund Live Stock Ins. v. Olson, 22 Pa. D. 437 (C.P. 1913) (“This policy of insurance is a contract made and to be performed in Williamsport, Lycoming County, Pennsylvania.”).
  135. . Gleason v. Nw. Mut. Life Ins., 203 N.Y. 507, 510 (1911).
  136. . Fountain v. Sec. Mut. Life Ins., 93 S.E. 118, 119 (Ga. 1917).
  137. . Hay et al., supra note 76, at 19–24.
  138. . Id. at 65–75.
  139. . Id. at 67.
  140. . Id. at 72.
  141. . Id. at 58–63.
  142. . Restatement (Second) of Conflict of Laws § 188(1) (Am. Law Inst. 1971) (emphasis added).
  143. . It is important to note that clauses that contain language keyed to vested rights have always comprised a relatively small percentage of all choice-of-law clauses. It is much more common for such clauses to utilize the words “interpret,” “construe,” or “govern” than to explicitly reference a specific place, as discussed in Sections II.2 and II.3.
  144. . William F. Elliott, Commentaries on the Law of Contracts 774 (1913).
  145. . Id.
  146. . Dwight Arven Jones, Construction or Interpretation of Commercial and Trade Contracts 3 (1886) (emphasis added).
  147. . Elliott, supra note 143, at 774.
  148. . Louis L. Hammond, General Principles of the Law of Contract 775 (1902).
  149. . Joseph Story, Commentaries on the Conflict of Laws 232 (1834).
  150. . Wharton, supra note 33, at 884.
  151. . Beale, supra note 68, at 1201–05.
  152. . The drafters of the Second Restatement took the position that the courts must always apply the law of the forum to interpret a contract. See Restatement (Second) of Conflict of Laws § 204 cmt. a (Am. Law Inst. 1971). If the meaning of the contract is ambiguous, however, then the courts should seek to determine which law should be used in order to construe the agreement. Where the contract contains a choice-of-law clause, the court should then apply the law chosen by the parties to construe the contract. Id. § 204 cmt. b. Section 204 has been cited twelve times since it was published. In none of these cases did the courts distinguish between the act of “interpreting” a contract and the act of “construing” it.
  153. . See infra note 153 (collecting sources); cf. H. G. Craig & Co. v. Uncas Paper Bd. Co., 133 A. 673, 675 (Conn. 1926) (“The appellant . . . contends that the word ‘made’ should be construed as ‘governed.’ We are unable to adopt the latter contention; the only meaning of which ‘made’ as here used is fairly susceptible is as relating to the place of signing and delivery and as thus fixing the locus contractus as being in Pennsylvania.”).
  154. . See Boat Town U.S.A., Inc. v. Mercury Marine Div. of Brunswick Corp., 364 So. 2d 15, 17 (Fla. Dist. Ct. App. 1978) (“In the instant case, there is no assertion, nor could any be substantiated, that ambiguities exist in the terms of the contract. Thus, the interpretation clause of the contract has no effect and does not provide an explicit choice of Wisconsin law to govern the conduct of the parties.”); see also Arnone v. Aetna Life Ins., 860 F.3d 97, 107–08 (2d Cir. 2017) (“In the context presented here, that provision [stating that the agreement will be ‘construed’ in accordance with Connecticut law] is insufficient to bind this court to apply the full breadth of Connecticut law, to the exclusion of another jurisdiction’s law, in fields other than the interpretation of the language in this contract.”); Am.’s Favorite Chicken Co. v. Cajun Enters., 130 F.3d 180, 182 (5th Cir. 1997) (“Since the . . . claims do not implicate the interpretation or construction of the franchise agreements, they are not governed by the narrow choice of law clause present here.”); Dollar Sys., Inc. v. Avcar Leasing Sys., Inc., 890 F.2d 165, 171 (9th Cir. 1989) (observing that the plaintiff’s “compliance with franchise law did not depend on the construction of the license agreement” and that the choice-of-law clause was therefore inapplicable); Procter v. Mavis, 125 P.3d 801, 803 (Or. 2005) (“The choice-of-laws provision in the parties’ premarital agreement is just that: an agreement that California law will govern the construction of the agreement. The provision does not relate to the law applicable to the division of property on dissolution.” (emphasis in original)). But see Hammel v. Ziegler Fin. Corp., 334 N.W.2d 913, 916 (Wis. Ct. App. 1983) (“We . . . can conceive of few instances where it would be reasonable to look to the law of a specific state to define contractual terms but to the law of a second jurisdiction to ascertain the legal effect of the agreement. Such a maneuver would be unreasonable because the meaning associated with a term by one jurisdiction might not mesh with the statutory and common-law scheme of another.”); see also John F. Coyle, The Canons of Construction for Choice-of-Law Clauses, 92 Wash. L. Rev. 631, 656–61 (2017) [hereinafter Coyle, Canons of Construction].
  155. . Boatland, Inc. v. Brunswick Corp., 558 F.2d 818 (6th Cir. 1977); C.A. May Marine Supply Co. v. Brunswick Co., 557 F.2d 1163 (5th Cir. 1977).
  156. . Knieriemen v. Bache Halsey Stuart Shields, Inc., 427 N.Y.S.2d 10, 12 (N.Y. App. Div. 1980).
  157. . Id.
  158. . Id. This outcome was consistent with those reached by other courts asked to decide whether a choice-of-law clause encompassed the tort law of the chosen jurisdiction in the late 1970s. See Travis v. Harris Corp., 565 F.2d 443, 446 (7th Cir. 1977) (“The parties and the district court agreed that Ohio law should be applied because the 1964 contract so provided. Though the contract may be interpreted under Ohio law, the legal effect of that agreement, and questions of traditional tort law unrelated to the contract, are to be determined in accord with the laws of Indiana, the situs of the injury and domicile of Travis.”); Indus. Consultants, Inc. v. HS Equities, Inc., 1980 U.S. Dist. LEXIS 10961, at *8 (S.D.N.Y. Apr. 14, 1980) (concluding choice-of-law clause stating that the agreement was to be “construed” in accordance with New York law did not cover action for misrepresentation); Barron v. Kane & Roach, Inc., 398 N.E.2d 244, 246 (Ill. 1979) (“The question before us . . . is unrelated to a construction of the provisions of the contract. That question of traditional tort law is to be determined in accordance with the law of Illinois, the situs of the injury and the domicile of the injured party.”).
  159. . See Peter A. Alces, Guerilla Terms, 56 Emory L.J. 1511, 1513 (2007) (observing that “it is not in rational form drafters’ interest” to bring [one-sided contract provisions] to the attention of less sophisticated consumers”); Robert Prentice, Contract-Based Defenses in Securities Fraud Litigation: A Behavioral Analysis, 2003 U. Ill. L. Rev. 337, 386 (2003) (“Farnsworth . . . has noted that in his own experience in legal practice, ‘no one in any of the corporations or in the law firm ever suggested that the forms should be drafted other than as one-sidedly in the interests of the corporate client as possible.’” (quoting E. Allan Farnsworth, On Trying to Keep One’s Promises: The Duty of Best Efforts in Contract Law, 46 U. Pitt. L. Rev. 1, 44 (1984))).
  160. . Salovaara v. Jackson Nat’l Life Ins., 66 F. Supp 2d 593, 596 (D.N.J. 1999) (“related . . . to”); Anderson v. First Commodity Corp., 618 F. Supp. 262, 264 (W.D. Wis. 1985) (“in connection with”).
  161. . Rooney v. Biomet, Inc., 63 F. Supp. 2d 126, 127 (D. Mass. 1999) (“relationship”); Am. Standard Leasing Co. v. Plant Specialties, Inc., 427 So. 2d 555, 556 (La. Ct. App. 1983) (“tort”).
  162. . See, e.g., Krock v. Lipsay, 97 F.3d 640, 645 (2d Cir. 1996) (“Under New York law, in order for a choice-of-law provision to apply to claims for tort arising incident to the contract, the express language of the provision must be ‘sufficiently broad’ as to encompass the entire relationship between the contracting parties.”).
  163. . It appears that the first two decades of the twenty-first century have witnessed a shift in this regard. A recent review of 351 choice-of-law clauses written into bond indentures filed with the SEC in 2016 found that approximately 12 percent contained language intended to give the clause a broader scope. John F. Coyle, Choice-of-Law Clauses in U.S. Bond Indentures, 13 Cap. Mkts. L.J. 152 (2018) [hereinafter Coyle, U.S. Bond Indentures]. A separate review of 159 international supply agreements filed with the SEC between 2011 and 2015 found that approximately 22 percent contained similar language. John F. Coyle & W. Mark C. Weidemaier, Interpreting Contracts Without Context, 67 Am. U. L. Rev. 1673 (2018).
  164. . Carl Hagenbeck & Great Wallace Show Co. v. Randall, 126 N.E. 501, 502 (Ind. App. 1920) (emphasis added).
  165. . See Michael Gruson, Governing Law Clauses Excluding Principles of Conflict of Laws, 37 Int’l L. 1023, 1025 (2003).
  166. . Duskin v. Pa.-Cent. Airlines Corp. 167 F.2d 727, 732 (6th Cir.), cert. denied, 335 U.S. 829 (1948).
  167. . See Siegelman v. Cunard White Star Ltd., 221 F.2d 189, 194–96 (2d Cir. 1955) (criticizing this approach); see also Coyle, Canons of Construction, supra note 153, at 642–47 (discussing the distinction between internal law and whole law).
  168. . Siegelman, 221 F.2d at 194.
  169. . Restatement (Second) of Conflict of Laws § 187(3) (Am. Law Inst. 1971); id. at cmt. h.
  170. . Id. § 187, cmt. h.
  171. . First Wis. Nat’l Bank v. Nicolaou, 270 N.W.2d 582, 585 (Wis. Ct. App. 1978) (“The proper construction of the choice of law provision . . . is that it makes Wisconsin internal law, i.e., the law which would govern purely domestic Wisconsin cases, applicable to any act of enforcement of the contract wherever it may occur.”).
  172. . Carlos v. Philips Bus. Sys., Inc., 556 F. Supp. 769, 774 (E.D.N.Y. 1983).
  173. . McLaughlin v. Reynolds, 886 F. Supp. 902, 904 (D. Me. 1995) (“without regard to its conflict of laws rules”).
  174. . Packquisition Corp. v. Packard Press New Eng., No. 91-5966, 1992 U.S. Dist. LEXIS 17460, at *7 (E.D. Pa. Nov 1, 1992).
  175. . First Nat’l Bank of Mount Dora v. Shawmut Bank of Bos., 389 N.E.2d 1002 (Mass. 1979); see also Wiesenberger Servs., Inc. v. Response Analysis Corp., 365 F. Supp. 258, 258 (S.D.N.Y. 1973) (“This agreement shall be governed by, and construed in accordance with the laws of the State of New York, other than conflicts of law rules.”).
  176. . This upward trend appears to have continued in the first two decades of the twentieth century. A review of 351 choice-of-law clauses written into bond indentures filed with the Securities Exchange Commission in 2016 found that 55 percent contained language excluding conflict-of-laws rules. Coyle, U.S. Bond Indentures, supra note 162. A separate review of 159 international supply agreements filed with the SEC between 2011 and 2015 found that approximately 78 percent contained language excluding conflict-of-laws rules. Coyle & Weidemaier, supra note 162.
  177. . 600 Grant St. Assocs. Ltd. v. Leon-Dielmann Inv. P’ship, 681 F. Supp. 1062, 1064 (S.D.N.Y. 1988).
  178. . Brinderson-Newberg Joint Venture v. Pac. Erectors, Inc., 690 F. Supp. 891, 892 (C.D. Cal. 1988) (“Except as provided in the General Contract as incorporated herein, this subcontract shall be interpreted and governed substantively and procedurally, including periods for limitations of actions, by the law of the State of California.”).
  179. . Clauses specifically referencing statutes of limitations are not included on the chart because there were only two such clauses in the data set.
  180. . It is not altogether clear that drafters are fully cognizant of the significance of the word “enforced” when they write it into their clauses, which in turn makes it hard to assess the significance of clauses that use this term. This fact notwithstanding, a number of courts have held that the use of the word “enforced” in a choice-of-law clause connotes an intent to select the procedural law of the chosen jurisdiction. See Coyle, Canons of Construction, supra note 153, at 655 n.113 (collecting cases).
  181. . Stark v. Nw. Nat’l. Life Ins., 167 F. 191, 192 (D. Minn. 1909) (contract stating that “the laws of the state of Minnesota . . . constitute and form a part of each . . . policy”).
  182. . See Cotter v. Lyft, Inc., 60 F. Supp. 3d 1059, 1066 (N.D. Cal. 2014); Steven W. Feldman, Statutes and Rules of Law as Implied Contract Terms: The Divergent Approaches and a Proposed Solution, 19 U. Pa. J. Bus. L. 809, 852 (2017).
  183. . See Alaska Packers Ass’n v. Indus. Accident Comm’n, 1 Cal. 2d 250 (1934) (“[T]he parties . . . hereby agree, as an incident of this contract of employment, to accept and be bound by the provisions of the said Workmen’s Compensation Act of Alaska for any and all injuries arising out of and in the course of their employment, and further agree to accept as their exclusive remedy for any and all industrial injuries, the provisions of the said Workmen’s Compensation Act of Alaska.”).
  184. . Only a smattering of clauses expressly referenced a federal treaty. Vision Graphics, Inc. v. E.I. Du Pont de Nemours & Co., 41 F. Supp. 2d 93, 97 (D. Mass. 1999) (“THIS AGREEMENT WILL BE GOVERNED BY AND CONSTRUED IN ACCORDANCE WITH THE LAWS OF THE STATE OF DELAWARE without regard to its choice of law provisions (and the United Nations Convention on the International Sale of Goods shall specifically not apply).”).
  185. . U.S. Fid. & Guar. Co. v. Nw. Eng’g Co., 112 So. 580, 581 (Miss. 1927) (involving a contract stipulating that the rights of the parties were to be “governed by the laws of Wisconsin existing at the time of making the contract”).
  186. . Stark, 167 F. at 192 (involving a contract stipulating that it was to be governed by “the laws of the state of Minnesota as they now exist or as may hereafter be amended”).
  187. . See Vitek Danilowicz, ‘Floating’ Choice-of-Law Clauses and Their Enforceability, 20 Int’l Law. 1005 (1986).
  188. . See English Co. v. Nw. Envirocon, 663 N.E.2d 448, 451 (Ill. App. Ct. 1996) (“The Agreement shall be tried or heard in and governed by and construed in accordance with the governing laws of the home state of the initial defendant to any action at law brought by or on behalf of the other party, including countersuit by the initial defendant.”).
  189. . Dean Witter Reynolds, Inc. v. Espada, 959 F. Supp. 73, 75 (D.P.R. 1997) (“The law of the State of New York will apply in all respects, including but not limited to determining of applicable statutes of limitation and available remedies.”).
  190. . Cf. Espresso Disposition Corp. 1 v. Santana Sales & Mktg. Grp., Inc., 105 So. 3d 592, 594 (Fla. Dist. Ct. App. 2013).
  191. . See Michael I. Meyerson, The Efficient Consumer Form Contract: Law and Economics Meets the Real World, 24 Ga. L. Rev. 583, 605 (1990) (“Intuitively, any profit-maximizing business would prefer to shift a risk to the other party if it could do so at no additional cost. Because consumers lack the knowledge to evaluate the cost of the risk, a rational seller will draft contract terms that shift risks to the consumer.”).
  192. . See Robert A. Hillman & Jeffrey J. Rachlinski, Standard-Form Contracting in the Electronic Age, 77 N.Y.U. L. Rev. 429, 452 (2002) (“This narrow cognitive focus that people bring to complex decisions creates a temptation for businesses to offer enticing prices and terms concerning the negotiable portions of the form and to make up for any concessions by drafting one-sided boilerplate terms. Consumers will focus their cognitive skills on the ‘important’ terms, such as price, but ignore the hidden costs buried in the boilerplate.”).
  193. . Russell Korobkin, Bounded Rationality, Standard Form Contracts, and Unconscionability, 70 U. Chi. L. Rev. 1203, 1244 (2003) (“[A]lthough the market should be expected to provide efficient salient contract terms to the advantage of buyers as a class and sellers as a class, no such assumption about non-salient terms is defensible.”).
  194. . In one sense, it may be rational for some lawyers to invest relatively little time in updating and revising their choice-of-law clauses. Clients are generally uninterested in paying their attorneys to research and revise contract boilerplate, and attorneys (rationally) tend to devote their time to substantive contract issues that are of greater concern to their clients.
  195. . Even if these companies were concerned about whether these clauses would be enforced, there was no penalty for including them in the agreement. At worst, the judge would ignore the clause and apply the law required under the forum’s conflict rules.
  196. . Path dependence likely goes a long way towards explaining the stickiness of contract language. In the absence of any pressing reason to update a contract, the path of least resistance is to leave it be. This is particularly true when the provision at issue is of marginal interest to nonlawyers and is typically found at the very end of the contract along with other “miscellaneous” provisions.
  197. . See Wendy Netter Epstein, Public-Private Contracting and the Reciprocity Norm, 64 Am. U. L. Rev. 1, 35 (2014) (observing that “the positive reciprocity norm—meaning that people reward kind actions—has been shown to often constrain actors’ behavior, resulting in deviations from what the rational actor model would predict”).
  198. . Stephen Choi, Mitu Gulati & Eric A. Posner, The Dynamics of Contract Evolution, 88 N.Y.U. L. Rev. 1, 9–10 (2013) [hereinafter Choi et al., Dynamics]; see also Stephen J. Choi, Mitu Gulati & Eric Posner, The Evolution of Contractual Terms in Sovereign Bonds, 4 J. of Legal Analysis 131, 152 (2012) (discussing the evolution of other clauses in sovereign debt agreements).
  199. . Choi et al., Dynamics, supra note 197, at 27 (“[S]hifts in boilerplate contract terms do not occur without some initial shock.”).
  200. . Florencia Marotta-Wurgler & Robert Taylor, Set in Stone? Change and Innovation in Consumer Standard-Form Contracts, 88 N.Y.U. L. Rev. 240, 247–48 (2013) (describing shocks as “changes in legal interpretations of terms, or technological advances”).
  201. . Choi et al, Dynamics, supra note 197, at 37 (“[T]op market participants switch from being defenders of the status quo to promoters of their own individual visions of the anticipated new standard.”).
  202. . See Stephen J. Choi & Mitu Gulati, Contract as Statute, 104 Mich. L. Rev. 1129, 1133, 1157 (2006) (“[W]ith boilerplate clauses, dispersed market participants may lack the ability to coordinate, at least initially, to clarify the language in subsequently adopted terms.”).
  203. . Mitu Gulati & Robert E. Scott, The Three and a Half Minute Transaction: Boilerplate and the Limits of Contract Design 60 fig.2 (2013).
  204. . Id. at 60 figs. 2 & 3.
  205. . Christopher Drahozal & Peter Rutledge, “Sticky” Arbitration Clauses? The Use of Arbitration Clauses After Concepcion and Amex, 67 Vand. L. Rev. 955, 982–83 (2014) (invoking shock model to explain changes to arbitration clauses).
  206. . Matthew Jennejohn, The Architecture of Contract Innovation, 59 B.C. L. Rev. 71, 89 n.83 (2018).
  207. . David Hoffman, Whither Bespoke Procedure?, 2014 U. Ill. L. Rev. 389, 394 (2014).
  208. . With respect to the pace of change, the shock model presupposes that the process of contractual change begins with something akin to a lightning bolt. In the choice-of-law clause context, however, a better metaphor is that of a pebble perched on the top of the mountain. If the pebble falls just the right way, it will dislodge additional pebbles, which will in turn dislodge still more pebbles, until the avalanche begins in earnest.
  209. . This list of search terms included (1) “choice of law provision,” (2) “choice of law clause,” (3) “in accordance with the laws of the state” w/30 contract, (4) “construed in accordance with the laws of the state,” (5) “governed by” w/20 “according to the,” (6) “construed” w/5 “according to the laws,” (7) “governed by and construed,” (8) “the laws of the state of” w/20 “provision,” (9) “the laws of the state of” w/20 “interpreted,” (10) “the law of the state” w/20 “interpreted,” (11) “construed” w/5 “interpreted,” (12) “construed” w/5 “governed by,” (13) “interpreted” w/5 “governed by,” (14) “construed solely,” (15) “determined in accordance with the laws” w/30 contract, (16) “determined in accordance with the laws” w/30 “agreement,” (17) “in accordance with the laws of” w/30 “agreement,” (18) “governed” w/5 “interpreted,” (19) “governed” w/5 “construed,” (20) “construed” w/5 “interpreted,” (21) “shall be deemed to have been made,” (22) validity w/10 “shall be determined,” (23) “shall be governed as to validity,” (24) “the place of the contract,” (25) agreement w/5 “subject to,” (26) “Without regard to” w/5 conflict!, (27) construction w/10 interpretation w/10 agreement, (28) construction w/10 validity w/10 agreement, (29) “this agreement and its enforcement,” (30) “agreement” w/8 “deemed” w/8 “made,” and (31) “agreement” w/8 “deemed” w/8 “executed.”
  210. . See, e.g., David A. Hoffman, Alan J. Izenman & Jeffrey R. Lidicker, Docketology, District Courts, and Doctrine, 85 Wash. U. L. Rev. 681 (2007); Ahmed E. Taha, Data and Selection Bias: A Case Study, 75 UMKC L. Rev. 171 (2006).

In the field of conflict of laws, private actors are generally granted the power to choose the law to govern their contracts. This is the doctrine of party autonomy. In recent years, this doctrine has been the subject of several excellent histories that draw upon judicial opinions, scholarly writings, and legislative enactments to chronicle changing attitudes toward party autonomy over time. A moment’s reflection, however, reveals that judges, scholars, and legislatures are not the most important actors in this story. The true protagonists are the contracting parties who write choice-of-law clauses into their agreements, without which there would be no need for any doctrine of party autonomy. These drafters and their creations are, however, almost entirely absent from the existing histories.

This Article seeks to remedy this deficit. It provides answers to certain basic questions about choice-of-law clauses that cannot be found in the existing literature. When did they first appear? Have they always been popular? Has the manner in which they are drafted changed over time? It describes how these provisions were first used in the years immediately after the Civil War by companies operating in a small number of industries. It shows how they slowly found their way into a growing number of agreements in the early twentieth century before enjoying a “breakthrough” moment in the early 1960s. And it recounts how contract drafters in the late twentieth century began experimenting with new language that simultaneously expanded the reach of these clauses and prompted the courts to devise new interpretive rules.

This historical account, while interesting in its own right, also has broader implications. First, it underscores the extent to which contract drafters do not always function as rational actors. In some cases, drafters added language to their clauses that was arguably unnecessary. In others, they declined to add language that would have advanced their interests more effectively. Second, it shows that the pattern of contractual change over time in the context of choice-of-law clauses is different than the pattern observed with respect to other types of contractual provisions. This finding suggests the need for new models of contract innovation.

Introduction 1149

I. Prevalence and Enforcement 1153

A. Party Intent and Implied Choice-of-Law 1154

B. The Express Choice-of-Law Clause 1156

C. The Judicial Response 1159

D. Expansion into New Types of Agreements 1161

E. Joseph Beale and the First Restatement 1165

F. From Novelty to Mainstream 1170

G. The Second Restatement and State Statutes 1175

1. Statutes Restricting Party Autonomy 1177

2. Statutes Expanding Party Autonomy 1179

H. The Choice-of-Law Clause in the Twenty-First Century 1181

II. Drafting and Interpretation 1182

A. Use of “Made,” “Executed,” “Performed,” and “Place” 1184

B. Use of “Interpret” and “Construe” 1187

C. Use of “Govern” 1189

D. Noncontractual Law 1193

E. Clauses that Exclude the Conflicts Rules of the Chosen Jurisdiction 1196

F. Procedural Law 1200

G. Other Language 1201

III. Implications for Contract Scholarship 1203

A. The Myth of the Rational Contract Drafter 1203

B. New Models of Contractual Innovation 1207

Conclusion 1210

Appendix: Published Case Data Collection 1211

Introduction

When a lawsuit has a connection to more than one jurisdiction, a judge called upon to resolve the dispute must perform a conflict-of-laws analysis to determine which jurisdiction’s law should be applied. In theory, this analysis is straightforward. The judge reviews the facts of the case, determines which state has the closest connection to the parties and the transaction, and applies the law of that state.[2] In practice, things are rarely so neat. Practitioners have long bemoaned the fact that case outcomes in this area can be difficult to predict.[3] And legal scholars have long argued that judges prefer to apply the (more familiar) law of their home jurisdictions and will sometimes use the flexibility inherent in conflicts doctrine to achieve this goal.[4]

Party autonomy—the idea that the parties to a contract may select in advance the law that will govern their contract—offers a solution to both of these problems.[5] When a contract contains a choice-of-law clause, it is easier to predict the outcome of a conflicts analysis because the court will typically apply the law chosen by the parties. The use of such clauses also curtails the ability of judges to engineer the selection of the law of their home jurisdictions. Perhaps for these reasons, choice-of-law clauses have become much more common over time. One recent study found that 75 percent of material contracts executed by public companies contain such a clause.[6] Having obtained the autonomy to select the law to govern their dispute, in short, it appears that many companies are taking full advantage.

The popularity of choice-of-law clauses in contemporary contract practice raises several questions. When did these clauses first appear? Have they always been popular? Has the manner in which they are drafted changed over time? Surprisingly, the existing literature provides few answers to these questions.[7] While the literature contains a number of excellent histories devoted to the concept of party autonomy, these works generally focus on judicial decisions, legislative acts, and scholarly writings.[8] They devote virtually no attention to the history of the contractual instruments by which contracting parties wield their autonomy.

This Article aspires to provide such a history. It shows that the doctrine of party autonomy rests on a contractual foundation. At step one, the parties draft a contract that contains a choice-of-law clause. At step two, the courts interpret the clause and determine whether it is enforceable. At step three, a legislature decides whether to intervene to codify or overturn the judicial decision. At step four, legal scholars discuss and evaluate these judicial decisions and legislative actions. And then the entire cycle begins again as contract drafters draft new choice-of-law clauses that take the foregoing developments into account. Absent the choice-of-law clause, there is no need for any doctrine of party autonomy.[9] And yet contemporary histories of the doctrine generally ignore these contractual provisions.

This oversight is likely attributable to the availability of sources. The doctrinal history of party autonomy is one that can be told through case decisions, scholarly treatises, and legislative enactments. All of these sources can be found on the shelves of a well-resourced law library. The history of the choice-of-law clause, by comparison, is a history that can only be told through a careful review of actual contracts. These agreements are not readily accessible to scholars. They are tucked away in aging file cabinets or stored in boxes in dusty warehouses. In order to study such clauses, one must be able to observe them. And the overwhelming majority of private agreements from long ago cannot easily be observed.

In an attempt to overcome these challenges—and to integrate the choice-of-law clause into the broader history of party autonomy—this Article looks to two original sources.[10] First, it draws upon more than two dozen contract form books published between 1860 and 2019 that contain thousands of “off-the-rack” contracts that were and are used by practicing lawyers to draft agreements for their clients. Second, it draws upon an original, hand-collected data set of more than three thousand choice-of-law clauses that appear in published cases decided between 1900 and 2000. This data set, which was assembled over a period of three years by a team of more than a dozen research assistants, represents the single largest historical collection of choice-of-law clauses currently in existence. Collectively, these sources reveal—for the first time—how contract drafters responded to the multitude of cases, legislative enactments, and treatises that grappled with the challenges posed by choice-of-law clauses beginning in the nineteenth century. In so doing, they make it possible to write the first meaningful history of the choice-of-law clause.

To illustrate the potential utility of such a history, consider the First Restatement of Conflict of Laws, published in 1934. This work took the position that a contract should be governed only by the law of the place of performance or the place of making. It made no reference to the concept of party autonomy and, in so doing, implicitly rejected that concept. The Second Restatement of Conflict of Laws, published in 1971, cast aside the position taken by the First Restatement and expressly endorsed the concept of party autonomy. As a matter of intellectual history, the shift from the First Restatement to the Second Restatement was hugely consequential. As a matter of drafting practice, however, nobody has the slightest idea whether the position taken by the First Restatement had any impact on the willingness of private actors to write choice-of-law clauses into their contracts between 1934 and 1971. It is possible that this seminal moment in the intellectual history of party autonomy had a major impact on drafting practice. It is also possible that it had no impact at all. We simply don’t know.

This Article seeks to answer this question (among others) by looking at choice-of-law clauses through the lens of contract rather than the lens of conflict of laws. This does not mean that conflict of laws is irrelevant to the present inquiry. It is impossible to understand many of the changes to the choice-of-law clause over the past century without taking conflicts doctrine into account. It is merely to point out that the clause itself—which is distinct from the concept of party autonomy—is a worthy object of study that has, at least to date, attracted little attention from legal scholars.

This Article is divided into three parts. Part I focuses on the prevalence of choice-of-law clauses and the debates among judges, legislators, and scholars as to whether these provisions are enforceable. It identifies the first express choice-of-law clauses in the late nineteenth century and traces the slow process by which these clauses came to be used in certain agreements in the years following the Civil War. It explains how these clauses forced the courts to develop an entirely new set of doctrinal rules to determine when such clauses should and should not be given effect. These judicial decisions, in turn, led to these clauses being written into new types of contracts, which prompted an academic debate about the underlying merits of party autonomy as a legal concept. The resolution of this debate in favor of party autonomy led to more choice-of-law clauses and, ultimately, legislative intervention. Part I concludes by discussing the role that these clauses play in modern contract practice.

Part II focuses on the interplay between contract drafters and the courts with respect to how to draft and interpret choice-of-law clauses. In some respects, the language in the standard choice-of-law clause has evolved a great deal over the past 150 years. In other respects, it has remained surprisingly static. The Part shows how each and every word in the modern choice-of-law clause has a specific meaning and, importantly, that this meaning may not be immediately obvious to the casual reader. It then tracks how certain words and phrases that commonly appear in these clauses have changed—or not—in response to judicial decisions over time.

Part III considers the implications of the foregoing analysis for contract law generally. It first argues that the history of the choice-of-law clause provides good reason to question whether contract drafters are, in fact, the rational actors assumed by contract theory. On a number of occasions, drafters have added language to these clauses that is arguably unnecessary. On other occasions, they have declined to add language that would arguably further their interests. The Part then argues that the leading explanation for how and why contract language evolves over time is a poor fit when applied to the choice-of-law clause. It can take decades for innovations to percolate across the contract landscape when it comes to choice-of-law clauses. This slow pattern of change stands in stark contrast to prior studies suggesting that contractual innovation is characterized by rapid and convulsive change. This difference suggests that the process of contractual evolution may vary depending on the precise nature of the provision at issue.

  1. Prevalence and Enforcement

This Part discusses the origins of the choice-of-law clause. It traces the clause’s slow rise to prominence in the nineteenth and twentieth centuries and shows how its growing popularity prompted spirited discussions among judges, scholars, and legislators as to whether and to what extent such provisions should be enforceable. Although the intensity of these debates may seem curious to a modern reader, the question of whether contracting parties had the power to choose the law to govern their agreements was hotly debated in the first half of the twentieth century. For these debates to occur, contracting parties had to start writing choice-of-law clauses into their agreements. This Part recounts the story of that first clause and everything that came after.

  1. Party Intent and Implied Choice-of-Law

In the nineteenth century, courts in the United States generally applied one of two tests to determine the law that would govern a contract.[11] The first test inquired into where the contract was made.[12] The second test inquired into where the contract was to be performed.[13] On occasion, however, the courts would nod to a third test. This test focused on the intent of the parties rather than the place where the contract was made or the place of performance.[14] Today, the notion that party intent should determine what law to apply is generally known as the doctrine of party autonomy.

The origins of party autonomy in Anglo-American law can be traced back to 1760. In that year, Lord Mansfield observed in the English case of Robinson v. Bland that “[t]he law of the place can never be the rule, where the transaction is entered into with an express view to the law of another country, as the rule by which it is to be governed.”[15] This position was subsequently adopted by the U.S. Supreme Court. In 1825, the Court stated in Wayman v. Southard that “universal law [recognizes] the principle that, in every forum, a contract is governed by the law with a view to which it was made.”[16] While express choice-of-law clauses were still largely unknown in this era, the notion that the courts should consider the implied intent of the parties in deciding which law to apply in a given dispute was not.[17]

The earliest reported choice-of-law clauses in the United States were, oddly, interest rate provisions.[18] Consider Butters v. Olds, a case decided in 1860. In this case, the Iowa Supreme Court was called upon to determine the law that would govern a loan agreement. Although the agreement was made in Iowa, the interest was to be paid to the lender in New York. The legal rate of interest in New York was 7 percent. The legal rate of interest in Iowa was 10 percent. The contract called for the borrower to pay interest at a rate of 10 percent. If the contract was governed by the law of the place of performance (New York), it would be invalid for failure to comply with that state’s usury laws. If the contract was governed by the place where the contract was made (Iowa), it would be valid and enforceable.

In concluding that the loan agreement was governed by the laws of Iowa, the Iowa Supreme Court held that the parties had, by selecting the rate of interest permitted under Iowa law, chosen that law to govern their agreement. In its words:

If the rate of interest in the place of the contract differs from that in the State where it is to be performed, it is competent for the parties to stipulate for the rate of interest in either locality, and thus by their own contract determine which law shall govern this incident thereof. In this case the parties did expressly stipulate by their contract for the rate of interest allowed by our statute . . . .[19]

The interest rate provisions in this and similar contracts laid the foundation for the modern choice-of-law clause. In these cases, the courts relied on specific contract language—rather than a holistic analysis of implied intent based on the surrounding circumstances—to hold that the contract would be governed by the law of a particular jurisdiction. While this analytical approach was initially used with respect to a particular class of agreements (loan agreements) and a particular body of law (usury laws), it paved the way for creation of the express choice-of-law clause. Once the principle that the parties could choose a particular state’s usury laws by writing a specific interest rate into their contract was established, it logically followed they should also be able to choose the contract law of that same jurisdiction by including a statement to that effect in their agreement.

  1. The Express Choice-of-Law Clause

The earliest known express choice-of-law clause in the United States appears in a lending agreement executed in 1869.[20] That clause provided that the contract was “made under, and is in all respects to be construed, by the laws of the state of Illinois.”[21] In the decades that followed the Civil War, choice-of-law clauses were gradually introduced into standard-form contracts in industries that did extensive business across state lines.

As life insurance companies began offering policies on a nationwide basis in the 1880s and 1890s, they increasingly utilized choice-of-law clauses to mitigate the legal risks associated with doing business in many different states. In 1882, the Union Central Life Insurance Company entered into an insurance contract with a Virginia resident which stipulated that it was to “be held and construed to have been made in the city of Cincinnati, Ohio.”[22] In 1887, the New York Life Insurance Company issued a policy to a Minnesota resident which stated that “the entire contract contained in the said policy and in this application . . . shall be construed and interpreted . . . according to the laws of the State of New York.”[23] And in 1892, the Mutual Benefit Life Insurance Company issued a policy to an Iowa resident which stipulated that it “shall at all times and places be held and construed to have been made in the city of Newark, New Jersey.”[24] In choosing the law of its home jurisdiction to govern these agreements, each insurer sought to ensure that its contracts would be governed by a single, uniform law, even as it sold policies in many different states.

International transportation companies, which were dominated by foreign interests throughout this era, were also early adopters of express choice-of-law clauses. In 1885, an English company entered into a shipping contract with the Baltimore-based agent of a Chicago businessman to transport cattle from Baltimore to Liverpool. This contract stated that “[a]ny questions arising under this contract or the bill of lading against the steamer or her owners shall be determined by English law in England.”[25] In 1897, a Belgian steamship company sold a ticket to two passengers who wished to travel from Antwerp to New York City.[26] This ticket provided that “all questions arising hereunder are to be settled according to the Belgium law, with reference to which this contract is made.”[27] The purpose of such clauses, again, was to minimize the legal uncertainty that invariably accompanies the act of transporting goods and people from one nation to another.

Late nineteenth-century mortgage lenders doing business across state lines also wrote choice-of-law clauses into many of their contracts. In 1881, for example, the New England Mortgage Security Company entered into a mortgage agreement with an Oregon borrower which was to “be construed according to the laws of Oregon, where the same is made.”[28] In 1886, a Missouri borrower entered into a loan agreement with a New York lender stipulating that “this bond and the interest notes hereto annexed are made and executed under and are in all respects to be governed and construed by the laws of the state of Missouri.”[29] And in 1893, a Minnesota borrower issued a promissory note to a Connecticut lender that contained the following clause: “It is expressly agreed and declared that these notes are made and executed under, and are in all respects to be construed by, the laws of the state of Minnesota.”[30]

It is something of a mystery as to why mortgage lenders typically chose the law of the borrower’s jurisdiction as opposed to the law of the lender’s jurisdiction. In all likelihood, the lenders believed that any litigation arising out of the transaction would occur in the borrower’s home jurisdiction. The odds that the borrower would default, after all, were far greater than the odds that the lender would refuse to make the loan, and it was unlikely that the borrower would be subject to personal jurisdiction outside of his home jurisdiction. Once litigation began, moreover, the courts in the borrower’s home jurisdiction were likely to apply the law of that jurisdiction because that was where the real property was located. If this hypothesis is correct, then the purpose of the choice-of-law clause in mortgage lending agreements was not to give the lender an advantage in litigation or even to ensure a uniform law would be applied to lending agreements concluded in many different states. It was to reduce costs by eliminating any uncertainty as to the governing law at the time of litigation.

In any event, the historical record is clear that express choice-of-law clauses were regularly written into life insurance policies, transportation contracts, and mortgage agreements at the close of the nineteenth century. Inevitably, some of these clauses wound up in litigation before the courts. The reaction of those courts is described below.

  1. The Judicial Response

As express choice-of-law clauses proliferated, the courts were increasingly called upon to determine whether and to what extent such clauses should be enforced. Citing Joseph Story’s acclaimed treatise on conflict of laws, Phillimore’s commentaries on international law, and the decision by the U.S. Supreme Court in Wayman v. Southard—all of which endorsed the general principle of party autonomy—U.S. courts in the late nineteenth century frequently concluded that these clauses were enforceable.[31] In a series of cases in the 1890s, the state supreme courts in Arkansas, Georgia, Minnesota, New York, and Virginia were presented with choice-of-law clauses that sought to displace traditional conflicts rules relating to contracts.[32] In each instance, the court concluded that the parties’ intent—as expressed in their choice-of-law clause—should determine the governing law.

This did not mean, of course, that the courts enforced every choice-of-law clause that came before them.[33] In 1882, a federal court in Missouri was called upon to decide whether a choice-of-law clause selecting the laws of New York in a life insurance contract should be given effect to the exclusion of a Missouri statute designed to protect local policyholders. In concluding that the choice-of-law clause was not enforceable, the court stated:

The statutes of Missouri, for salutary reasons, permit foreign corporations to do business in the state on prescribed conditions. If, despite such conditions, they can by the insertion of clauses in their policy withdraw themselves from the limitations of the Missouri statutes, while obtaining all the advantages of its license, then a foreign corporation can by special contract upset the statutes of the state and become exempt from the positive requirements of law. Such a proposition is not to be countenanced. The defendant corporation chose to embark in business within this state under the terms and conditions named in the statute. It could not by paper contrivances, however specious, withdraw itself from the operation of the laws, by the force of which it could alone do business within the state.[34]

In other cases, the courts cited public policy as a rationale for declining to enforce a choice-of-law clause. In 1893, the U.S. District Court for the Southern District of New York was asked to decide whether a clause selecting the law of England in a bill of lading operated to discharge the carrier from responsibilities imposed by U.S. law. In concluding that it could not, the court observed that:

The objections to the validity of stipulations exempting common carriers from responsibility for negligence, namely, the policy of the law of this country, the unequal situation of the parties, and the lack of sufficient evidence of actual intention and freedom of contract, apply precisely the same to a stipulation for the adoption of the law of another country, as to the original exemption. That [choice-of-law clause selecting English law] is plainly nothing but a further device to secure the same unlawful exemption as the preceding exemption clause, which could not stand alone.

Such an additional stipulation, so far as it relates to the same exemption of liability for negligence, must fall with the latter. Nor can a rule of law founded on public policy be set aside in our own court by any stipulation to adopt the law of another country.[35]

In summary, while many U.S. courts in the late nineteenth century endorsed the notion that parties could select the law to govern their contracts, they also recognized that this autonomy was subject to limits. In particular, these courts took the position that parties could not use a choice-of-law clause to evade the mandatory law of a jurisdiction with a close connection to the parties. They also held that these clauses were unenforceable when contrary to public policy. These limitations notwithstanding, the courts were willing to give effect to these clauses in at least some cases. These decisions, unsurprisingly, led to choice-of-law clauses being written into new types of contracts.

  1. Expansion into New Types of Agreements

In the wake of judicial decisions recognizing choice-of-law clauses as generally enforceable, these clauses gradually came to be incorporated into an ever-wider range of agreements. In 1914, an English company entered into an agreement to acquire substantially all the assets of a company headquartered in Washington state.[36] In 1917, an Illinois-based manufacturer entered into an agreement with an Arkansas-based dealer to sell motor trucks.[37] In 1918, a Chicago-based company entered into a contract to sell heavy machinery to an Indiana-based company.[38] Although each case involved a different type of contract—an asset purchase agreement, an exclusive distribution agreement, and a sales agreement, respectively—each of these contracts contained a choice-of-law clause.

These clauses also began to find their way into employment agreements, particularly when the employee’s work required extensive travel. In 1918, a circus company and one of its employees entered into a contract in which each party acknowledged that the employee’s “performance under this contract shall embrace services and travel in and through the several and various states of the U.S.A. and the Dominion of Canada.”[39] In light of this fact, the parties agreed “that the place of the contract and release, its status or forum, is the District of Columbia” and that “all matters whether sounding or in tort, relating to its validity, construction, and interpretation [shall] be determined to the same extent as if its execution, performance, or cause of action thereon or growing out of the same, actually took place or arose in said District of Columbia.”[40]

The burgeoning motion-picture industry also made extensive use of choice-of-law clauses.[41] In the early 1920s, motion-picture distributors based in New York leased motion pictures to exhibitors across the United States. These agreements frequently contained a choice-of-law clause selecting New York law to govern the provisions of the contract relating to arbitration.[42]

Choice-of-law clauses also appeared in a number of noncommercial agreements during this era.[43] In 1912, a husband living in Virginia sought a divorce from his wife—who was then living in New York—on grounds of desertion.[44] In connection with the divorce proceeding, the parties entered into a separation agreement which stated that it “should be construed according to the laws of Virginia.”[45] In 1914, a wife sought to obtain a divorce from her husband in New York on the grounds of adultery.[46] In connection with this proceeding, the parties entered into an agreement that addressed such questions as alimony, counsel fees, and custody of the couple’s only child.[47] This contract stipulated that it was “to be construed according to the law of the State of New York.”[48]

For ease of reference, the earliest example of a published case referencing a choice-of-law clause in a particular type of contract is set forth in Table 1. This information was gleaned from reported cases rather than actual contracts. Accordingly, it is possible—indeed, it is likely—that the first actual use of a choice-of-law clause in each type of agreement was earlier than the one listed below. Nevertheless, the Table provides a general sense for the slow process by which choice-of-law clauses found their way into different types of agreements before and after the turn of the twentieth century.

This expansion notwithstanding, it is important to emphasize that choice-of-law clauses were not a standard feature of most contracts in the first decades of the twentieth century.[49] Although some companies made regular use of these clauses in certain contexts, many others did not.[50] The cases discussing those clauses that did exist, however, soon began to attract attention from legal scholars. One of these scholars was Joseph Beale.

Table 1. Earliest Choice-of-Law Clause Referenced in Reported Case, by Contract Type.
Contract Type Year of Contract Contract Type Year of Contract
Bond[51] 1869 Distribution Agreement[52] 1910
Mortgage[53] 1872 Asset Purchase Agreement[54] 1914
Prenuptial Agreement[55] 1874 Motion Picture Licensing Agreement[56] 1917
Life Insurance Policy[57] 1880 Sales Agreement[58] 1918
Bill of Lading[59] 1885 Stock Purchase Agreement[60] 1918
Steamship Ticket[61] 1897 Employment Agreement[62] 1918
Trust[63] 1904 Contract of Bailment[64] 1922
Separation Agreement[65] 1909 Rental Agreement[66] 1930
  1. Joseph Beale and the First Restatement

Joseph Beale was a professor at Harvard Law School, the founding dean of the University of Chicago Law School, the reporter for the First Restatement, and the author of an acclaimed multivolume treatise on conflict of laws.[67] In the early decades of the twentieth century, Beale used his position as the leading conflicts scholar in the United States to wage a war against party autonomy and choice-of-law clauses. It was a war that he ultimately lost. In order to understand why, it is important to discuss Beale’s reasons for waging the war in the first place.

Beale’s opposition to the principle of party autonomy stemmed from three sources. First, he believed that the idea that the parties should be permitted to choose the law to govern their agreement was “alien” to the common law tradition.[68] While the continental states of Europe had long endorsed the concept of party autonomy, he argued, it had never been fully embraced by the Anglo-American legal system. In support of this argument, he pointed out that Lord Mansfield had relied upon continental writers when he endorsed the concept in 1760.[69] This was, in Beale’s view, grounds for concern: “Is it then permissible for us to base principles of the Conflict of Laws on civil-law authorities? It is submitted in general that this should not be done.”[70]

Second, Beale argued that the U.S. Supreme Court had been less than steadfast in its commitment to the principle of party autonomy in the years since the Court first endorsed it in 1825.[71] In support of this argument, he identified a number of more recent conflicts cases in which the Court had articulated a conflicts rule relating to contracts—to apply the law of the place where the contract was made, for example—without making any mention of party intent. The persuasiveness of Beale’s analysis in this regard is undercut by the fact that the issue of party intent was apparently not raised as an issue in many of the cases that he cited. Nevertheless, he invoked the Court’s failure to mention party intent in every conflicts case that came before it as proof that the Court was growing increasingly hostile to the principle of party autonomy.[72]

Third, Beale argued that party autonomy was inconsistent with the vested rights theory of conflict of laws. This theory posited that all legal rights vested in a specific time and place and thereafter traveled with the plaintiff wherever he went.[73] For example, if a person were injured in an accident in Mississippi and subsequently brought a lawsuit against the tortfeasor in Alabama, the suit would be governed by Mississippi law because the right to bring a tort claim had “vested” in Mississippi at the moment of the injury.[74] The proponents of the vested rights theory believed that each state was sovereign within its own borders and that no state could make a law that would operate in any other state.[75] It logically followed from this proposition that the parties to a contract were powerless to select the law that would govern their agreements. As Beale explained:

The fundamental objection to [party autonomy] in point of theory is that it involves permission to the parties to do a legislative act. It practically makes a legislative body of any two persons who choose to get together and contract . . . and gives to the parties what is in truth the power of legislation so far as their agreement is concerned.[76]

If all lawmaking power was vested in the state, and if each state lacked the power to legislate outside of its territorial borders, then it was irrelevant whether a contract contained an express choice-of-law clause. The only legal rights that could possibly have been invoked by a plaintiff were those that had been vested by the state with sovereign control over the territory where the cause of action arose.

If Beale had simply made these arguments in the context of his own academic writings, they would have been influential enough; Beale towered over the field of conflicts law for more than fifty years.[77] As mentioned above, however, Beale also served as the reporter for the First Restatement of Conflict of Laws. In this capacity, he was tasked with “restating” the common law of conflicts. When it came time to draft those sections of the First Restatement dealing with the issue of party autonomy, Beale put forward his own vision of what the law should be rather than restating the law as it was.[78] Notwithstanding the extensive caselaw endorsing the principle of party autonomy discussed above, the First Restatement contains not a single reference to it. Indeed, the First Restatement is devoid of any hint of the possibility that party intent has a role to play in conflict of laws.[79]

Shortly after the First Restatement was published in 1934, Beale made an observation and a prediction. He observed that the “tendency [among the courts] is to abandon the law intended by the parties as the law to govern the validity of a contract.”[80] He then predicted that it was “probable that before many years have passed the influence of the American Law Institute will have led to the abandonment of the doctrine of intention of the parties.”[81] As it turns out, Beale was incorrect on both counts.

In an article published in 1942, Arthur Nussbaum took issue with Beale’s “observation” that U.S. courts were becoming less and less likely to enforce choice-of-law clauses.[82] Nussbaum noted that the intent theory was one of the most “time-honored” and “universally adopted” rules of private international law.[83] He observed that the U.S. Supreme Court had endorsed the general principle of party autonomy on numerous occasions and had commented favorably on the use of choice-of-law clauses. Nussbaum also identified a number of methodological problems with respect to Beale’s case analysis purporting to show a trend away from enforcing choice-of-law clauses.[84] After canvassing the cases that Beale had relied upon to support his observation, Nussbaum concluded that they provided little to no support for the empirical claim that U.S. courts were becoming less likely to enforce choice-of-law clauses.[85] Beale’s descriptive account of recent judicial practice, in Nussbaum’s assessment, was an extended exercise in wishful thinking.

Beale’s prediction as to the future also proved inaccurate.[86] In 1937, just three years after the publication of the First Restatement, the U.S. Supreme Court doubled down on its prior decision in Wayman v. Southard when it observed that a choice-of-law clause in a contract between a Pennsylvania oil company and a Connecticut insurer was “undoubtedly” enforceable because “[i]n every forum a contract is governed by the law with a view to which it was made.”[87] In 1946, a panel of judges sitting on the Second Circuit made clear that they were familiar with the First Restatement and nevertheless believed that the parties should have the autonomy to select the law to govern their agreement:

Under orthodox conflict of laws rules this provision [for rescission] is to be given effect by the forum according to the law of the place of performance. Restatement, Conflict of Laws (1934) Secs. 358, 359. Where, however, there is no single place of performance, and there are many jurisdictions with which the contract has close association, there seems no reason, apart from the policy of the forum, why the parties cannot specify the law of one jurisdiction as controlling, so long as there is that sufficient relationship to make it reasonable that the law chosen should apply. Decisions to the seeming contrary have been so on the basis of the policy of the forum or on the ground that there was an apparent place of performance the law of which forbade choice of another law. Since no fixed policy to the contrary has been established in this court, [the choice-of-law clause selecting New York law] is controlling and the rights of the parties herein are to be determined by the law of New York.[88]

The final blow came in 1952, when the Uniform Commercial Code (UCC) was published.[89] The drafters of the UCC included a provision that expressly permitted the contracting parties to choose the law of any state that bore a “reasonable relation” to the contract.[90] At this point, the intellectual debate shifted away from whether the intent of the parties should be considered to whether party autonomy should be subject to any limits. This development, unsurprisingly, led to yet another uptick in the number of contracts that contained choice-of-law clauses, as recounted below.

  1. From Novelty to Mainstream

It is difficult to determine the impact of the foregoing developments—the publication of the First Restatement, the ensuing academic debates about the nature and scope of party autonomy, and the publication of the UCC—on the prevalence of choice-of-law clauses in U.S. contracts. This is because there is no easy way to estimate, as a historical matter, what percentage of contracts contained a choice-of-law clause in the early to mid-twentieth century. To calculate this number accurately, it is necessary to determine, first, the total number of contracts executed in a given year and, second, how many of these contracts contained choice-of-law clauses. From the vantage point of 2020, it is simply not possible to gather this information.

It is possible, however, to obtain a rough sense for the prevalence of such clauses by looking to books of form contracts that were published during the relevant time periods.[91] In an era before photocopiers—let alone computers and word processors—lawyers would routinely consult form books containing samples of many different types of contracts when called upon to draft a particular type of agreement.[92] General form books typically contained hundreds of agreements, organized by type, that could be quickly and cost-effectively deployed by the contract drafter when the need arose.[93] Since these books provide a historical record of what provisions were typically included for specific types of agreements at a particular point in time, they offer a means by which scholars can get a general sense for the prevalence of the choice-of-law clause in a particular era. One need only select a well-known form book from a given year, count the number of contracts in the book, and calculate the percentage of those contracts containing choice-of-law clauses. As a judge on the Pennsylvania Court of Common Pleas put it in 1930: “An examination of the form books, while not being recognized as stating the law, gives some indication of what the general practice is.”[94]

With this insight in mind, I reviewed twenty-five form books with the aid of two research assistants. The earliest form book dated to 1860. The contracts in that book contained not a single choice-of-law clause. The most recent form book dated to 2019. Sixty-nine percent of the contracts in that book contained a choice-of-law clause. The bulk of our time and attention was spent on form books published between these years. With respect to each book, we recorded the total number of contracts contained therein as well as the number of those contracts that contained a choice-of-law clause.

There are, of course, methodological problems with using the form books in this way. On the one hand, form books could be a leading indicator of actual party practice. If this is the case, then changes in the form books may precede changes in real-world contract drafting. On the other hand, form books could also be a trailing indicator of actual party practice. If this is the case, then changes in the form books may lag behind changes in real-world contract drafting. It is also possible that the form books represent a stylized version of contract practice that in no way reflects actual contracting practice during the era. If this is the case, then the form books may present a portrait of contract practice that affirmatively misleads the reader. Nevertheless, the form books arguably represent the best sources available to a researcher trying to ascertain the frequency with which choice-of-law clauses were incorporated into contracts drafted more than a century ago.

While looking to form books is an act of desperation, there is some reason to believe that these books offer a reasonably good guide to actual contract practice when it comes to choice-of-law clauses. A review of a modern form book available through Lexis in 2019—Forms from Rabkin & Johnson, Current Legal Forms with Tax Analysis—turned up 1,381 contracts. Nine hundred fifty-two of these agreements—69 percent—contained choice-of-law clauses. A recent empirical study that reviewed half a million actual contracts filed with the U.S. Securities and Exchange Commission (SEC) between 2000 and 2016 found that 75 percent of these agreements contained choice-of-law clauses.[95] The fact that these two percentages are in the same general ballpark—69 percent versus 75 percent—provides some comfort that form books from the past might offer a reasonably accurate measure of the prevalence of choice-of-law clauses in earlier eras.

A review of contract form books published between 1910 and 1965 suggests that although the choice-of-law clause was widely adopted by companies in a few select industries early on, such clauses were not particularly common prior to the early 1960s. Consider Sayler’s American Form Book, published in 1913, and Tiffany’s Legal and Business Forms, published in 1915.[96] Together, these two books contain 471 form agreements. Only two of these agreements include a choice-of-law clause (0.4 percent of the total). A review of three form books published in the 1920s tells a similar story. Collectively, these form books contained 901 form agreements.[97] Only thirteen of these agreements contained choice-of-law clauses (1.5 percent). In 1938, West published a massive, multivolume collection titled Modern Legal Forms that purported to reflect current trends in contract drafting.[98] A review of two volumes from this collection turned up 263 contracts but only six choice-of-law clauses (2.3 percent). Looking ahead to the 1940s, a form book published in 1946—Gordon’s Modern Annotated Forms of Agreement—contained 431 contracts but only eleven choice-of-law clauses (2.6 percent).[99] The 1946 edition of Jones Legal Forms Annotated tells a similar story.[100] Although it contained 678 contracts, only three of those contracts contained choice-of-law clauses (0.4 percent).

The fact that choice-of-law clauses were rarely written into contracts in form books published in the 1920s, 1930s, and 1940s suggests that Beale’s decision to omit any discussion of party autonomy from the First Restatement was of relatively little consequence from a drafting perspective. While it is possible that the position laid down by the First Restatement discouraged parties who would otherwise have written such clauses into their agreements from doing so—an empirical question that is extremely difficult to answer—it does not appear that this work prompted contract drafters to remove existing provisions from their agreements. The evidence from the form books suggests that, in most cases, there was nothing to remove.[101]

Not until the 1950s did drafting practice undergo a meaningful shift. While virtually all of the contracts in the fourteen-volume American Jurisprudence Legal Forms Annotated (1953) lacked choice-of-law clauses, the inside front cover of each volume contained a “checklist” of things to remember when drafting forms.[102] The third item on the checklist encouraged contract drafters to ask: “Should there be a declaration that the laws of a specific jurisdiction will govern interpretation, validity, performance, etc.?” The form books from the early 1960s are the first to contain a meaningful number of contracts with choice-of-law clauses. Approximately 21 percent of the 140 contracts in Forms of Business Agreements with Tax Ideas (1961) contained a choice-of-law clause.[103] And 19 percent of the 118 contracts in Forms for Commercial Transactions Under the Uniform Commercial Code (1963) contained such a clause. This marked a significant shift in practice, as shown in the table below.[104]

A screenshot of a cell phone Description automatically generated

It appears that the early 1960s represented something of a watershed moment for the choice-of-law clause. Although such clauses had been around for almost a century by this point, and although they had long been embraced by companies in a few select industries, there is no evidence that these clauses were written into contracts outside of these industries with any sort of regularity until the early 1960s. Indeed, I was unable to find a single form book published prior to 1961 where more than 3 percent of the contracts contained choice-of-law clauses. While the clause was known to prior generations of contract drafters, it was not widely used until the early 1960s.

There are many possible explanations for why the choice-of-law clause gained traction at this particular moment in history.[105] Due to space limitations—and the uncertain reliability of the underlying sources—only one possible explanation will be addressed here. This is the possibility that the enactment of the UCC spurred more parties to write choice-of-law clauses into their agreements. As discussed above, the draft UCC contained a provision that specifically directed courts to enforce choice-of-law clauses in commercial contracts when the chosen state bore a “reasonable relation” to the contract. Although the UCC was first published in 1952, it was substantially revised in 1956 and was not enacted by most states until the early 1960s.[106] It may not be a coincidence that there was an uptick in the number of choice-of-law clauses appearing in form books at the same moment when many states were in the process of enacting a statute that specifically authorized courts to enforce choice-of-law clauses in commercial agreements.[107] To be sure, this account is speculative. There is no easy way to determine why a widely dispersed body of private actors began to make changes to their private agreements many decades ago. It is, however, an explanation that is at the very least consistent with the patterns of practice suggested by the form books.

  1. The Second Restatement and State Statutes

The Second Restatement, which was finalized in 1969 and published in 1971, staked out a position on the issue of party autonomy that was very different from that of the First Restatement. With respect to default legal rules, the Second Restatement stipulated that the parties could choose the law of any state on the theory that they could have simply written the rule into their contract.[108] With respect to mandatory legal rules, the Second Restatement took the position that parties could select the law of a particular state to govern their agreement but that this ability was subject to two limits. The first limit requires that either the chosen law have a substantial relationship to the parties or the transaction or there be a reasonable basis for the parties’ choice.[109] The second limit requires that the chosen law not be contrary to a “fundamental policy of a state which has a materially greater interest . . . in the determination of the particular issue and which . . . would be the state of the applicable law in the absence of an effective choice of law by the parties.”[110]

Each of these limitations can be easily traced to prior legislative or judicial practice. The “substantial relationship” requirement bears a close resemblance to the “reasonable relation” requirement that appears in the UCC.[111] The “reasonable basis” requirement was derived from cases upholding the validity of clauses selecting the law of a significant commercial jurisdiction such as England or New York as a “neutral” law.[112] And the “fundamental policy” requirement can be traced back to cases from the late nineteenth century in which U.S. courts refused to uphold choice-of-law clauses in the face of local statutes requiring the application of forum law to certain types of disputes.[113] Unlike the First Restatement, the Second Restatement actually sought to restate the law as it was, as opposed to how its drafters wanted it to be. Perhaps as a result, section 187 is now among the most widely followed provisions in the Second Restatement.[114]

The enforceability of choice-of-law clauses is not, however, always determined by common law rules of the sort laid down by the Second Restatement. As choice-of-law clauses proliferated in the decades after 1960, many state legislatures chose to enact statutes that preempt these common law rules. Some of these enactments directed state courts to apply the law of the forum even when the parties had expressly selected the law of another jurisdiction. These statutes served to restrict party autonomy. Other enactments directed state courts to enforce choice-of-law clauses even when the law chosen lacked a substantial relationship to the parties or the transaction. These laws operated to expand party autonomy. Both types of statutes are discussed below.

Statutes Restricting Party Autonomy

State statutes that direct state courts to apply the law of the forum even when the parties have selected the law of another state have a long history. In the 1890s, Missouri enacted a statute designed to protect its residents against so-called “forfeiture” provisions in life insurance contracts.[115] The statute made clear that such provisions could not be enforced even when the contract in question contained a choice-of-law clause selecting the law of a state other than Missouri.[116] In this way, the state legislature restricted the ability of parties to wield their autonomy to select the law of another state.

In the second half of the twentieth century, states enacted other statutes restricting party autonomy with respect to specific types of contracts. The Uniform Consumer Credit Code, approved in 1968, directs courts to refuse to enforce choice-of-law clauses selecting the law of a state other than that of the borrower’s residence.[117] In the 1970s and 1980s, many states enacted franchise acts that directed state courts not to enforce any contract provision—including a choice-of-law clause—that would amount to a “waiver” of a right guaranteed to the franchisee under the statute.[118] Other states enacted similar anti-waiver provisions with respect to consumer contracts and employment contracts.[119] The overall effect of such provisions was (and is) to limit the ability of lenders, franchisors, manufacturers, and employers to utilize choice-of-law clauses to select the law that will govern their contractual relationships with borrowers, franchisees, consumers, and employees.[120]

While the foregoing statutes restricted party autonomy to address concerns about unequal bargaining power, other statutes restricted party autonomy based on a different set of concerns. In the first two decades of the twenty-first century, a number of state legislatures passed laws declaring choice-of-law clauses selecting non-U.S. law to be unenforceable under certain circumstances. The principal target of such statutes was Islamic law (or sharia), which was perceived by elected representatives in these states to pose a threat to American values.[121] The statutes in question generally direct state courts not to give effect to choice-of-law clauses selecting foreign law when to do so would abridge rights conferred upon the state’s residents under the U.S. Constitution or the state constitution.[122] The relevance of these statutes to real-world disputes is unclear—to date, it appears that such laws have actually been used to invalidate a choice-of-law clause selecting foreign law on only one occasion.[123] Nevertheless, such statutes represent a noteworthy legislative intervention designed to limit the autonomy of contracting parties to choose the law to govern their agreement.

Statutes Expanding Party Autonomy

Over the past few decades, state legislatures in the United States have also enacted several pieces of legislation designed to expand the autonomy of contracting parties. Under the common law approach set forth in section 187 of the Second Restatement, the parties’ choice of law will generally be honored if the chosen law has a “substantial relationship to the parties or the transaction” or if there is a “reasonable basis” for their choice. In the 1980s, however, a number of states enacted statutes that relaxed these restrictions for high-dollar-value commercial contracts. In 1984, for example, New York enacted a statute directing its courts to enforce choice-of-law clauses selecting New York law in commercial contracts for more than $250,000 even when the parties and the transaction lacked a “reasonable relation” to New York.[124] The legislature was transparent about its motivation in passing this law—it hoped to divert legal business to New York and away from other jurisdictions, thereby generating more business for New York lawyers.[125] The practical effect of this statute was to encourage companies with no other connection to New York to select that state’s law to govern their agreements, without any concern that the choice-of-law clause would be invalidated for the lack of any “substantial relationship” to New York.

In the years that followed, a number of other states followed New York in requiring their courts to enforce choice-of-law clauses selecting their law even where the transaction lacked a substantial relationship to the state.[126] In every case, the statutes did not direct state courts to enforce choice-of-law clauses generally. Instead, the statutes directed state courts to enforce clauses contained in qualifying contracts if, and only if, the clause selected the law of the enacting state. In light of this requirement, it would seem these enactments—like the one pioneered by New York—are less concerned with principles of party autonomy than they are with generating business for in-state lawyers.

A statute enacted by North Carolina in 2017 goes even further. This statute stipulates that a choice-of-law clause selecting North Carolina law in a business contract is enforceable even when the parties and the transaction lack a “reasonable relation” to the state.[127] The statute then goes on to provide that the same result should be obtained even when the contract contained a provision that was “contrary to the fundamental policy of the jurisdiction whose law would apply in the absence of the parties’ choice of North Carolina law.”[128] The end result is a legal regime in which the North Carolina courts will apply that state’s law to any business contract selecting the law of North Carolina, even when the transaction lacks a reasonable relation to the state and even when its law is contrary to a fundamental policy of a jurisdiction with a closer connection to the dispute.

  1. The Choice-of-Law Clause in the Twenty-First Century

Today, choice-of-law clauses are everywhere. A recent study of every contract filed with the SEC between 1996 and 2012 found that 70 percent contained a choice-of-law clause.[129] A subsequent study looking at the same database covering the time period from 2000 to 2016 pegged the number at 75 percent.[130] When the focus is narrowed to international contracts—agreements between a U.S. company and a foreign counterparty—the percentage rises still higher. One study of international supply agreements filed with the SEC between 2011 and 2015 found that 99 percent contained a choice-of-law clause.[131] There is little question that these clauses have become standard features in commercial agreements over the past century.

This outcome represents the culmination of a long conversation between contract drafters on the one hand, and judges, scholars, and legislatures on the other. Contract drafters wrote the first express clauses into their agreements in the nineteenth century, thereby initiating a judicial conversation about the proper role of party autonomy in conflict of laws. Other contract drafters incorporated these clauses into an ever-wider range of agreements in the early twentieth century, triggering the scholarly revolt against party autonomy led by Joseph Beale. Still other contract drafters began to make more extensive use of these clauses in the early 1960s, thereby prompting state legislatures to pass new laws spelling out the proper scope of party autonomy in a range of contexts involving consumers, employees, insureds, and franchisees. The end result is a complex tapestry of contract provisions, common law rules, statutes, and judicial commentary that defines the scope of party autonomy in the modern era.

* * *

The foregoing discussion focused on the prevalence of choice-of-law clauses and the conversations among judges, scholars, and legislatures as to whether and to what extent they should be enforceable. This is not, however, the only topic of historical interest relating to choice-of-law clauses. Another important question is whether the language in these clauses has changed in any meaningful way over the past 150 years. This issue is taken up below.

Drafting and Interpretation

The first known modern choice-of-law clause, which appeared in 1869, provided that the contract was “made under, and is in all respects to be construed, by the laws of the state of Illinois.”[132] In the decades that followed, thousands—and eventually millions—of choice-of-law clauses were written into other legal agreements. Inevitably, the language in these clauses changed and evolved. Certain words and phrases fell out of favor and were replaced. New legal issues arose that necessitated the addition of new language. These changes did not occur overnight. Indeed, the process of change was, in many cases, slow and halting. Over the course of decades, however, the typical choice-of-law clause gradually came to exhibit certain features that were distinct and different from those very first clauses drafted at the close of the nineteenth century.

This Part provides a historical account of these changes. As a prelude to this account, however, it is necessary to discuss questions of research methodology and data collection. It is obviously impossible to review and inspect every choice-of-law clause used in the tens of millions of U.S. contracts that entered into force over the past 150 years. In order to overcome these challenges, I turned to a somewhat unusual source: published cases. Over a period of several years, I worked with more than a dozen research assistants to comb through such cases in search of choice-of-law clauses. Whenever we found a clause referenced in a case, we inputted that clause—along with the year the contract containing the clause was signed and the type of contract at issue—into a spreadsheet. When our work was complete, we had collected 3,104 choice-of-law clauses that selected the law of a U.S. jurisdiction.[133] A comprehensive discussion of the data collection methods used to assemble this data set—along with several caveats to which its use is subject—is set forth in Appendix A. A table listing the number of clauses from various decades is set forth below.

Table 2. Number of Choice-of-Law Clauses Collected from Published Cases, by Contract Decade.
Decade

Contract Signed

Number of

Clauses

1880s 15
1890s 51
1900s 44
1910s 31
1920s 43
1930s 53
1940s 79
1950s 123
1960s 236
1970s 398
1980s 1020
1990s 1011
Total 3104

This Part explores the changing language in choice-of-law clauses across a number of dimensions. First, it looks to see whether certain words and phrases associated with the First Restatement—such as clauses that specifically reference the “place” of the contract—have become more or less common over time. Second, it discusses the relative frequency of the words “interpret” and “construe.” Third, it looks at the changing role of the word “govern.” Fourth, it seeks to determine whether “broad” choice-of-law clauses—those that select the tort and statutory law of the chosen jurisdiction in addition to its contract law—have become more common. Fifth, it discusses the rise of clauses that exclude the conflicts rules of the chosen jurisdiction. Sixth, it looks at whether clauses ever select the procedural law of the chosen jurisdiction. Finally, it briefly discusses additional words and phrases that may be of interest to scholars but that rarely appear in actual choice-of-law clauses.

Use of “Made,” “Executed,” “Performed,” and “Place”

Under the traditional vested-rights approach to conflict of laws, a contract was governed either by the law of the place where it was made or the law of the place where it was to be performed. To determine where a contract was made, the courts would generally look to where the offer was accepted. To determine where a contract was to be performed, the courts would generally look to where the promise was to have been performed. As principles of party autonomy gained traction in conflict of laws, and as companies began drafting contract language to take advantage of this autonomy, the resulting clauses would sometimes make express reference to one or both of these rules.

In 1881, for example, an insurance company issued a life insurance policy stipulating that “[t]his policy is a contract, made and executed in the State of New York, and shall be construed only according to . . . the laws of that State.”[134] In 1902, a different insurance company issued a policy stating that it was “a contract made and to be performed in the state of Wisconsin, and shall be construed only according to . . . the laws of said State.”[135] In 1902, still another insurance company issued a policy providing that “the place of this contract is expressly agreed to be in the city of Binghampton [sic], N. Y.”[136] In each case, the clause in question was drafted in such a way as to preemptively answer the questions that insurance companies expected courts committed to a theory of vested rights to ask—where was the contract made and where was it to be performed?

The vested rights approach to conflict of laws dominated the landscape in the early twentieth century.[137] In the 1950s and 1960s, however, U.S. courts began to experiment with new and different approaches, a process of experimentation now commonly described as the “Conflicts Revolution.”[138] In the realm of contracts, the primary effect of this Revolution was to de-emphasize the place of making and the place of performance as factors in determining the governing law.[139] Instead, courts began to inquire into which jurisdiction had the closest connection to the parties and the transaction.[140] This new approach, which was first adopted by the New York Court of Appeals in the mid-1950s, was eventually incorporated into the Second Restatement.[141] Section 188 of that work provides that, in the absence of a choice-of-law clause, “the rights and duties of the parties with respect to an issue in contract are determined by the local law of the state which, with respect to that issue, has the most significant relationship to the transaction and the parties under the principles stated in § 6.”[142] Under this new test, the place where the contract was made and the place where the contract was to be performed were factors for the court to consider in deciding which law to apply. At the end of the day, however, these factors were now part of a broader holistic analysis keyed to the concept of the most significant relationship.

As the Conflicts Revolution gathered steam, one would perhaps expect to see a decrease in the proportion of choice-of-law clauses that contained words such as “made” or “executed” or “performed” or “place.” All of these words, after all, were originally included in these clauses to answer questions posed by a (now-outdated) system for determining the law that should be applied to a contract with a connection to more than one jurisdiction.[143] There is, however, scant evidence that the Conflicts Revolution had much (if any) impact on the way that choice-of-law clauses were drafted, as shown below.

A close up of a map Description automatically generated

On the one hand, certain words associated with traditional conflicts rules have become slightly less common over time. For example, the proportion of clauses that refer to the place where the contract was “executed” declined from 5 percent in the 1940s to 3 percent in the 1990s. And the proportion of clauses that refer to the “place” of the contract declined from 4 percent in the 1940s to 1 percent in the 1990s. Conversely, other words associated with traditional conflicts rules have become slightly more common over time. The proportion of clauses that refer to the place where the contract was “made” rose from 6 percent in the 1940s to 7 percent in the 1990s. And the proportion of clauses that contain some variant of the word “perform” rose from 10 percent in the 1940s to 12 percent in the 1990s. This suggests that while the Conflicts Revolution of the 1950s and 1960s may have upended choice-of-law doctrine across the United States, its impact on how choice-of-law clauses were drafted was negligible.

Use of “Interpret” and “Construe”

In reviewing many thousands of choice-of-law clauses, two words appear over and over again. These words are “interpret” and “construe.” Some clauses provide that they shall be “interpreted in accordance with the laws” of a particular state. Other clauses provide that they shall be “construed in accordance with the laws” of a particular state. In light of this variation in drafting practice, it is sensible to inquire whether these two words are synonyms. In order to answer this question, it is necessary to take a brief detour into the law of contracts as it existed in the late nineteenth century.

In theory, it is possible to draw a distinction between the act of “interpreting” a contract and the act of “construing” that same agreement. The word “interpretation” technically refers to the process judges use to determine the meaning of the words actually used in an agreement.[144] The word “construction,” by contrast, refers to the process judges use to determine the meaning of the contractual relationship taken as a whole.[145] Under this framework, a contract must be interpreted before it may be construed. If the words in the contract are unambiguous as an interpretive matter, there is no need to construe the contract. If the words are ambiguous, however, the court may then draw upon various canons of construction to help determine their meaning.

While the distinction between interpretation and construction may well be important as a matter of contract theory, courts and commentators long ago concluded that this particular distinction was of little practical significance. As the author of an 1886 treatise on commercial contracts explained:

Throughout the present volume interpretation and construction are used as synonymous words. Attempt, however, has been made to distinguish between them and interpretation employed to denote merely the process of ascertaining the meaning of the terms of the contract, and construction to describe the further process of bringing this meaning into consonance with established rules of law. But they are used interchangeably by the judges, in the digests, and by the large majority of text writers, and notwithstanding the high authority of those who have endeavored to enforce the distinction, no sufficient reason for its adoption is perceived.[146]

Other commentators writing in the late nineteenth and early twentieth centuries took a similar position. One observed that “[t]he terms ‘interpretation of a contract’ and ‘construction of a contract’ are usually used interchangeably.”[147] Another pointed out that

the rules of interpretation and construction are necessarily more or less blended, since, in determining the meaning of a word, we have sometimes to appeal to the context, just as, in determining the meaning of the context, we have to throw ourselves into the position of the author when he used the particular words.[148]

As a matter of contract law, therefore, the distinction between “interpretation” and “construction” was not viewed as particularly meaningful at the turn of the twentieth century.

This was also broadly true as a matter of conflict of laws. As early as 1834, for example, Justice Story wrote that “it is now a principle, generally received, that contracts are to be construed and interpreted according to the laws of the state, in which they are made, unless from their tenor it is perceived, that they were entered into with a view to the laws of some other state.”[149] In 1905, Francis Wharton stated much the same rule in his treatise without drawing any distinction between the two terms.[150] Even Joseph Beale, writing in 1934, used the words “interpret” and “construe” interchangeably when discussing the parties’ ability to select the law that was to govern their contract.[151] In the eyes of most conflict-of-laws commentators in this era, therefore, the words “interpret” and “construe” meant the same thing.[152]

A review of the clauses in the data set reveals that it is far more common for these clauses to utilize the word “construe” than the word “interpret,” as illustrated below.

A screenshot of a map Description automatically generated

Notwithstanding diligent searching, I have not found any cases where a court attached any significance to the parties’ decision to utilize the word “interpreted” rather than the word “construed” in a choice-of-law clause. While it is probably more advisable to use the word “construe” in light of the technical distinctions outlined above, the courts invariably treat the two words as synonymous. Perhaps as a consequence, the relative proportion of clauses using the words “interpret” and “construe” has changed very little over the past few decades.

Use of “Govern”

In deciding what law to apply to resolve a contract dispute, courts have traditionally distinguished between several issues. The first is whether the contract is valid, that is, whether the promises made by the parties in the contract are binding. As a general rule, courts will always apply the law of the forum—rather than any law chosen by the parties—to resolve this question because the issue of validity is a threshold issue that must be addressed to determine whether a contract exists in the first place. The second issue is whether the choice-of-law clause should be given effect. A court may conclude, for example, that the parties entered into a valid contract but that the choice-of-law clause should not be given effect because applying the chosen law would contravene some strong public policy of the forum. Whether a clause should be given effect is, again, an issue generally determined by the law of the forum.

Once a court has determined that the contract is valid and the choice-of-law clause should be enforced, it must then determine the breadth of the choice-of-law clause. Does the clause select all of the contract law of the chosen jurisdiction, including any rules relating to default terms, performance, defenses, breach, and remedies? Or does it only select that portion of the chosen jurisdiction’s contract law that specifically relates to the interpretation or construction of otherwise valid contracts?

In answering this question, some courts have attached great significance to the precise words used by the parties in their choice-of-law clause. These courts have held that the use of the word “govern” signals an intent to select all of the contract law of the chosen jurisdiction. The use of the words “interpret” or “construe,” by contrast, only signals an intent to select that portion of the chosen jurisdiction’s contract law that specifically relates to interpretation.[153] To be clear, most U.S. courts have not endorsed this interpretive rule. The majority of courts—reasoning that it would be curious indeed for the parties to select the interpretive law of the chosen jurisdiction while leaving unanswered what contract law should be applied to other issues—have held that “govern,” “interpret,” and “construe” are synonyms. Beginning in the late 1970s, however, a minority of U.S. courts took the position that the words “interpret” and “construe” signal the parties’ intent to select only that portion of the chosen jurisdiction’s contract law that specifically relates to the interpretation or construction of otherwise valid contracts.[154]

Although only a few case decisions ever formally endorsed this distinction, the relative proportion of choice-of-law clauses utilizing the word “govern” experienced a sharp uptick beginning in the 1970s. Between 1940 and 1970, the percentage of choice-of-law clauses that contained this word held relatively constant at around 40 percent. In the 1970s, however, the percentage jumped to 55 percent. In the 1980s, it surged to 68 percent. In the 1990s, it rose again to 73 percent. Significantly, this uptick occurred even as the prevalence of “interpret” and “construe” remained relatively constant.

 

A screenshot of a cell phone Description automatically generated

This shift in drafting practice cannot be wholly attributed to judicial decisions suggesting that the words “interpret” and “construe” were too narrow to encompass all of the chosen jurisdiction’s contract law. The first judicial decision to assign a narrow meaning to these words dates to 1978, which postdates the initial uptick in choice-of-law clauses containing the word “govern” by several years. It seems likely, however, that these judicial decisions assigning a broader scope to the word “govern” played some role in driving the increase in the proportion of clauses that contain that word in the 1980s and 1990s.

This uptick is particularly noteworthy because the caselaw assigning a narrow meaning to the words “interpret” and “construe” did not represent the majority rule. As discussed above, most U.S. courts that addressed this issue in the late 1970s and early 1980s concluded that “interpret,” “construe,” and “govern” were synonyms.[155] This fact notwithstanding, a significant number of contract drafters apparently chose to rewrite their choice-of-law clauses to include the word “govern” to guard against the risk that a court might assign their clause a narrow meaning.

Noncontractual Law

The discussion above considered whether a particular choice-of-law clause selected all of the chosen jurisdiction’s contract law or only the portion of that jurisdiction’s contract law relating to contract interpretation. This was not, however, the only scope question to come before the courts in the late 1970s. In other cases, the courts were asked to decide whether a given choice-of-law clause merely selected the contract law of the chosen jurisdiction or whether it selected the tort and statutory law of the chosen jurisdiction

In the mid-1970s, for example, an investor entered into a contract with a stockbroker in the city of New Orleans.[156] This contract contained a standard choice-of-law clause stipulating that “[t]his contract shall be governed by the laws of the State of New York.”[157] After the stockbroker lost all of the investor’s money, the investor brought suit against the stockbroker in New York state court for breach of contract, negligence, and fraud. The court held that while the law selected in the contract’s choice-of-law clause covered the breach of contract claim, it did not cover the plaintiff’s claims for negligence and fraud because the language in the clause did not sweep broadly enough to cover noncontractual claims. After performing a conflicts analysis, the court concluded that while the plaintiff’s contract claims were governed by the laws of New York, per the choice-of-law clause, his claims for negligence and fraud were governed by the law of Louisiana, per the court’s conflicts analysis.[158]

In the wake of this and similar decisions, one might expect to see an increase in the proportion of choice-of-law clauses that select the noncontractual law of the chosen jurisdiction. Lawyers tasked with drafting choice-of-law clauses on behalf of their corporate clients, after all, generally wish to reduce legal uncertainty to the greatest possible extent. And broad choice-of-law clauses make it easier to predict the law that will govern any noncontractual claims brought against the company in the future.[159] There is, however, no evidence that contracting parties across the United States responded to these decisions by writing broader choice-of-law clauses into their contracts in the 1980s and 1990s. Throughout this time period, parties continued to draft clauses that selected the contract law—and only the contract law—of the chosen jurisdiction.

To understand the basis for this conclusion, it is necessary first to understand that there are many different ways to draft a clause that selects the non-contract law of a given jurisdiction. A clause could state that the chosen law shall apply to all claims that “relate to” or have some “connection with” the contract.[160] Alternatively, a clause could select the “tort law” of the chosen jurisdiction or state that the “relationship” between the parties shall be governed by the law of a given state.[161] Even taking these different possible formulations into account, however, just 2 percent of the 1970s clauses, 1 percent of the 1980s clauses, and 2 percent of the 1990s clauses contain language selecting non-contract law.

A screenshot of a cell phone Description automatically generated

Given the relative speed with which the drafting community responded to the late-1970s decisions assigning a narrow ambit to the words “interpret” and “construe,” the lack of any sort of widespread response to the mid-1970s decisions concerning noncontractual law is surprising. The issue was not one of capacity—the courts were very clear that the parties had the power to prospectively select the tort and statutory law of a given jurisdiction.[162] Most companies simply chose not to revise their clauses to take advantage of this opportunity.[163] This inaction is all the more surprising in light of a clause from 1918 that I uncovered during my research. This clause was part of a standard employment contract for circus performers who traveled around the country. It provided that

it is mutually agreed between the parties hereto that the place of the contract . . . is the District of Columbia, [and in said District of Columbia] or according to the laws thereof, if construed or litigated elsewhere, shall all matters whether sounding [in contract] or in tort, relating to its validity, construction, and interpretation be determined to the same extent as if its execution, performance, or cause of action thereon or growing out of the same, actually took place or arose in said District of Columbia.[164]

Stumbling across this particular clause in a contract from 1918 was a bit like opening an ancient Egyptian tomb for the first time and discovering a functioning helicopter inside. The lawyers drafting contracts for this circus company knew enough to prospectively select the applicable tort law decades before other companies began dabbling with the practice in the 1960s and 1970s. The notion that a contract drafter could preemptively select the tort law of a given jurisdiction, in summary, should not have come as a surprise to attorneys in the closing decades of the twentieth century. Nevertheless, the proportion of contracts containing choice-of-law clauses that selected the noncontractual law of the chosen jurisdiction was negligible.

Clauses that Exclude the Conflicts Rules of the Chosen Jurisdiction

If attorneys in the United States were curiously apathetic to the possibilities presented by broad choice-of-law clauses, they were inordinately alarmed by the risks presented by clauses that did not expressly exclude the conflict-of-laws rules of the chosen jurisdiction. To understand these risks—and why they are unlikely to come to pass in practice—it is necessary to take a brief detour into conflict-of-laws doctrine.

Conflicts scholars frequently distinguish between the internal law of a given state, excluding its conflict-of-laws rules, and the whole law of a given state, including its conflict-of-laws rules.[165] This distinction has the potential to generate ambiguity as to what “laws” the parties are referencing in their choice-of-law clauses. Did they mean to select the chosen state’s internal law? Or did they mean to select its whole law? Since the purpose of the choice-of-law clause is to reduce legal uncertainty, it stands to reason that most parties intend to select the internal law of the chosen state. Nevertheless, many choice-of-law clauses do not clearly state which type of “laws” they are choosing.

In the 1940s and 1950s, U.S. courts reached different conclusions as to how this textual ambiguity should be resolved. In 1947, the Sixth Circuit concluded that a choice-of-law clause selecting the “laws” of Pennsylvania referred to the whole law of that state. After reviewing the relevant Pennsylvania conflicts rules, the court concluded that—per the choice-of-law clause—the contract was governed by the law of Alabama.[166] Needless to say, this outcome was probably not the one intended by the parties when they chose Pennsylvania law to govern their agreement.[167] In 1955, by contrast, the Second Circuit was called upon to determine the meaning of the word “laws” in a choice-of-law clause selecting the laws of England. The court concluded that the word “laws” should be read to select the internal law of England because the entire purpose of the clause was to ensure a uniform result regardless of where the litigation occurred.[168] If the clause were interpreted to select the whole law of England, the court reasoned, then this objective could not be obtained because it could potentially lead to the application of the laws of still another nation.

In the wake of these conflicting decisions about the core meaning of an increasingly popular contractual provision, contract drafters should have immediately set about revising their choice-of-law clauses to make clear their intent. If these revisions occurred, they are absent from the historical record. There is not a single clause from the 1950s or the 1960s that addresses the internal law‌/whole law distinction. The first contract that I found selecting the internal law of a state dates to 1970.

While there is no way to know precisely why such clauses first began appearing in the early 1970s, the most likely explanation is that they were responding to the Second Restatement, which was finalized in 1969 and published in 1971. That work made a point to emphasize that the word “law” in choice-of-law clauses should typically be read to refer to the internal law of the chosen state rather than its whole law.[169] In the comments to section 187, for example, the drafters observed that:

When they choose the state which is to furnish the law governing the validity of their contract, the parties almost certainly have the [internal law] rather than the [whole law] of that state in mind. To apply the [whole law] of the chosen state would introduce the uncertainties of choice of law into the proceedings and would serve to defeat the basic objectives, namely those of certainty and predictability, which the choice-of-law provision was designed to achieve.[170]

Even though the position taken by the Second Restatement accorded with the preferences of virtually all parties, and even though this position was followed by virtually every U.S. court to consider the issue, there was still the lingering possibility that a judge might misinterpret the intended meaning of the word “laws” in a choice-of-law clause.[171] To guard against this remote possibility, some contract drafters began to redraft their clauses.[172] In some cases, they selected the “internal” laws of the chosen jurisdiction. In others, they stated that the court was to apply the law of the chosen jurisdiction “without regard to principles of conflict of laws.”[173] In still others, they directed the courts to apply the law that would be “applicable to contracts made and to be performed” in the chosen jurisdiction.[174] Although the language used in each case was different, the impulse was the same—to ensure that the courts did not apply the conflicts rules of the chosen jurisdiction were the contract ever to wind up in litigation.

Although the practice of drafting clauses to exclude the conflicts rules of the chosen jurisdiction began around 1970, such clauses were rarely found in contracts executed over the course of the following decade. Only 1 percent of the clauses from the 1970s contain language excluding the conflicts rules of the chosen jurisdiction.[175] In the 1980s, however, 8 percent of the clauses excluded these rules. This number jumped again in the 1990s, where the issue was addressed in 18 percent of clauses.[176]

A screenshot of a cell phone Description automatically generated

It is something of a mystery why this language caught on in the 1980s. Contract drafters had, after all, failed to respond to a number of other events in prior decades that should, in theory, have triggered a change in drafting practice. Whatever the cause, the data show that, beginning around 1980, an increasing proportion of choice-of-law clauses contained language excluding the conflicts law of the chosen jurisdiction.

Procedural Law

As a general rule, courts have held that choice-of-law clauses operate to select the substantive law of the chosen jurisdiction but not its procedural law. In other words, a choice-of-law clause selecting New York law is deemed to select that state’s (substantive) contract law but not its (procedural) law of evidence. On occasion, however, a choice-of-law clause purports to select the procedural law of the chosen jurisdiction. Some clauses provide, for example, that the agreement shall be “enforced” in accordance with the chosen jurisdiction’s law.[177] Other clauses provide that the “remedies” available to the parties shall be determined by the chosen jurisdiction’s law. Still other clauses state outright that the courts should apply the “procedural law” of the jurisdiction named in the clause.[178] The figure below highlights the relative frequency of clauses containing such language.[179]

A screenshot of a map Description automatically generated

 

Clauses that select the “procedural law” of a given jurisdiction, or that state that the “remedies” available to the parties shall be determined in accordance with that law, are exceedingly rare. By comparison, there are a fair number of clauses that specify that the agreement shall be “enforced” in accordance with the laws of the chosen jurisdiction; the proportion of such clauses has ranged between 10 percent and 20 percent over the past several decades. Such provisions operate primarily to select certain aspects of the chosen jurisdiction’s law—like statutes of limitations—that straddle the line between substance and procedure and to direct the courts to apply these aspects regardless of how they are characterized.[180]

Other Language

In reviewing thousands of choice-of-law clauses, unusual or distinctive provisions inevitably appear. Although such clauses do not warrant extensive analysis, this Section briefly discusses several additional words and phrases that occasionally appear in choice-of-law clauses. Although these unusual or distinctive clauses present interesting conceptual issues, they rarely appeared in contracts executed between 1880 and 2000.

First, there were almost no clauses that expressly incorporated the law of a given state into the contract by reference.[181] While courts have long recognized that parties have the ability to incorporate such law into their agreements, the choice-of-law clauses in the data set rarely did so.[182] Second, there were virtually no clauses that referenced a specific state or federal statute.[183] In the overwhelming majority of cases, the parties simply selected the “law” or “laws” of the chosen jurisdiction without referencing individual statutes by name.[184] Third, the number of clauses that selected the law of a given state at a particular point in time was negligible. A few clauses specified that they were selecting the law as it existed at the time the contract was made.[185] A few other clauses specified that they were selecting the law as it existed at the time of litigation.[186] The overwhelming majority of clauses, however, did not address the issue at all. Fourth, there were very few “floating” choice-of-law clauses in the data set.[187] A floating choice-of-law clause stipulates that the contract shall be governed by the law of State A if suit is brought in State A but that it shall be governed by the law of State B if suit is brought in State B.[188] Such clauses constituted a tiny percentage of the clauses in the data set. Fifth, and finally, there were virtually no clauses that expressly referenced the chosen jurisdiction’s statutes of limitations.[189]

Implications for Contract Scholarship

Legal scholars have long recognized that contract language can be stubbornly resistant to change. Indeed, there is a thriving academic literature that seeks to explain why some contract language is sticky while other language is not. This Part seeks to integrate the insights from the foregoing discussion of choice-of-law clauses into this literature. It first discusses the pitfalls of trying to use the rational-actor model to explain changes to the choice-of-law clause over the years. It then shows how the best-known general theory of contractual change fails to provide a convincing account for the process of evolution and change in the choice-of-law clause context.

The Myth of the Rational Contract Drafter

The choice-of-law clause is a standard piece of boilerplate. As such, it rarely attracts much in the way of attention from the contracting parties.[190] When a contract is being negotiated by two individuals with unequal bargaining power, moreover, the choice-of-law clause is typically offered as a take-it-or-leave-it proposition. The stronger party—almost always the contract drafter—will demand that the agreement be governed by the law of its home jurisdiction. The weaker party must accede to this demand if it wants to complete the transaction. In such cases, a rational drafter has every incentive to draft a one-sided choice-of-law clause that favors its interests. In addition to selecting the law of its home jurisdiction, for example, the rational drafter would be well-advised to prepare a broad choice-of-law clause—one that selects the tort and statutory law of that jurisdiction—to further reinforce the advantages that flow from litigating with reference to a familiar body of law. If the other side is going to agree to the deal anyway, so the argument goes, a drafter would be foolish not to include a broad clause in the agreement.[191]

Past studies have discussed why it is rational for the weaker party to sign a contract that contains these one-sided provisions. Briefly, these studies have found that many contract terms are non-salient to the weaker party.[192] A consumer purchasing a new car, for example, is far more likely to focus on the price of the vehicle and on the warranties offered by the dealer than on the choice-of-law clause in the bill of sale. While this state of affairs may produce inefficient contract terms in the aggregate, it is arguably rational from the perspective of the weaker party.[193] The overwhelming majority of purchases, after all, do not result in a lawsuit or even the threat of a lawsuit.

While the existing literature provides a convincing explanation for why it is generally rational for weaker parties to accept one-sided contract provisions, that literature has surprisingly little to say about why stronger parties sometimes fail to draft provisions that maximize their advantages. If the weaker party is going to enter into the transaction regardless of whether the contract contains a choice-of-law clause—and regardless of whether that clause is narrow or broad—then it is arguably irrational for the contract drafter not to include a broad choice-of-law clause selecting the noncontractual law of its home jurisdiction. This is especially true when the stronger party is utilizing the same template for all transactions of a given type. Contract drafters do not, however, always behave rationally when drafting choice-of-law clauses.

Consider the question of whether to write a choice-of-law clause into a contract in the first place. In the 1880s, a great many insurance companies selling life insurance policies across state lines wrote choice-of-law clauses selecting the law of the insurer’s home jurisdiction into their agreements. This was an eminently rational decision. These clauses were non-salient to prospective policyholders but could provide the insurer with important tactical advantages should the contract ever be litigated. Over the next eighty years, however, most other contract drafters chose not to write these clauses into their agreements. It was not until the early 1960s that these clauses began their slow march into the mainstream. This pattern of practice is very difficult to explain.[194] Choice-of-law clauses simultaneously eliminate the need to perform a conflicts analysis in the event of a dispute and offer manufacturers and other corporate entities the opportunity to lock in the contract law of their home jurisdiction.[195] And yet these provisions were rarely used outside of a few industries in the first half of the twentieth century.

Precisely the same critique may be applied to these corporate entities’ collective failure to write broad choice-of-law clauses selecting noncontractual law into their standard form agreements. As discussed above, there are examples of clauses that select the tort law of a given jurisdiction as early as 1918. In the closing decades of the twentieth century, however, this broad language was virtually never found in choice-of-law clauses. This pattern of practice is exceptionally difficult to explain. It is one thing to add an entirely new provision to an agreement. It is quite another to add a short phrase—“claims related to the agreement”—to an already-existing contract clause when the addition would clearly advance the interests of the contract drafter. And yet this phrase was rarely added.

This omission becomes even more mystifying in light of the fact that many of these same companies did revise their choice-of-law clauses to exclude the conflict-of-laws rules of the chosen jurisdiction. As discussed above, the risk that a modern court would ever interpret the word “laws” in a choice-of-law clause to select the whole law of the chosen jurisdiction, including its conflicts rules, is miniscule. Such a decision would be contrary to mountains of precedent as well as the position laid down by the Second Restatement. And yet the number of clauses that specifically exclude the conflicts rules of the chosen jurisdiction rose significantly in the 1980s and 1990s. While there is certainly no harm in drafting clauses in this manner, it is not entirely clear why this particular innovation caught on while other potentially more useful innovations (like broad clauses selecting noncontractual law) did not. All of this, again, highlights the challenges of applying a rational-actor model to explain the process of contract drafting.

It is tempting to think of contract drafting as a modular process akin to an automobile assembly line, in which language is added or removed in a mechanical and systematic way. While it is understandable why this model appeals to people—it seems like this is how things should work—it does a poor job of explaining the changes to the choice-of-law clause described above. Contract drafters do not always behave rationally when it comes to these clauses. In some cases, they incorporate new language that is arguably unnecessary. In other cases, they omit language that would arguably serve to better protect their interests.[196] The foregoing history, in short, provides a useful corrective to the view that stronger contracting parties inevitably leverage their bargaining power to force weaker parties to accept provisions that favor the drafter. Whether out of kindness or ignorance, contract drafters do not always act in this way.[197]

New Models of Contractual Innovation

In recent years, a number of scholars have sought to develop a theoretical model that seeks to explain how and why contractual innovations occur. The most influential model is one put forth by Stephen Choi, Mitu Gulati, and Eric Posner to explain changes to a specific provision—the collective action clause (CAC)—in sovereign debt agreements.[198] This model is generally known as the “shock model.”

The shock model posits that the process of contractual innovation proceeds in three stages. In stage one, a particular standard form dominates the market. Stage one comes to an end when an external “shock” disrupts the status quo.[199] While the precise nature of this shock is highly variable—it can be legal, political, or economic—the shock serves as the catalyst for the process of change and kicks off stage two.[200] In stage two, marginal law firms begin to experiment with contractual innovations that depart from the dominant standard form. The effect of these actions is to reduce the dominance of the existing standard form. In stage three, elite law firms and other market leaders begin to promulgate their own innovations with the goal of establishing a new standard.[201] These innovations continue until practice gradually coalesces around a single proposal. This innovation then becomes the new standard form and the cycle begins again.

In the highly centralized world of sovereign debt agreements and collective action clauses, this model works. In the highly decentralized world of general contracts and choice-of-law clauses, it works less well.[202] Unlike collective action clauses, choice-of-law clauses appear in many different types of contracts. Unlike collective action clauses, choice-of-law clauses are frequently drafted by lawyers who do not work at large firms.[203] And unlike collective action clauses, choice-of-law clauses are often prepared by lawyers who do not work in New York City.[204] All of these differences call into question the explanatory force of the shock model as applied to contractual innovations outside of the context of sovereign debt agreements. In the clubby world of sovereign debt, it may be possible for a relatively small contracting community to coalesce around a single proposal and to quickly incorporate that language into agreements of that type going forward. In the decentralized world of non-sovereign-debt contracts, by comparison, it is difficult for contract drafters preparing different types of agreements for clients in different industries in different parts of the country to coalesce around much of anything.

To the extent that these disparate actors rally around a common standard, moreover, the process of change is likely to be slower than the one envisioned by the shock model. In the years since that model was first proposed, it has been invoked by a number of other contracts scholars to suggest that, as a general matter, the process of contractual change is rapid and explosive.[205] Matthew Jennejohn, for example, has argued that “boilerplate evolves according to a punctuated equilibrium model of institutional change, by which concentrated and often dramatic adjustments follow long stretches of stasis.”[206] And David Hoffman has invoked the model to argue that “[c]ontract terms do not evolve linearly and progressively in rational counterpoint to slow changes in doctrine. They change contingently, explosively, and at moments punctuated by shocks.”[207]

While the process of contractual change as it relates to collective action clauses may well be quick and dramatic, contractual change proceeds differently in the context of choice-of-law clauses. Between 1930 and 1970, for example, the percentage of such clauses that contained some variation of the word “govern” held relatively constant at around 40 percent. In the 1970s, the percentage rose to 55 percent. In the 1980s, it rose to 68 percent. In the 1990s, it rose again to 73 percent. The process of change was not “explosive” or punctuated by “shocks.” Instead, it was slow, steady, and incremental.[208] The increase in the proportion of choice-of-law clauses that excluded the conflicts rules of the chosen jurisdiction—which went from 1 percent in the 1970s to 8 percent in the 1980s to 18 percent in the 1990s—was similarly incremental. In some cases, it would seem, the process of contractual innovation is more akin to the tide coming in than to a tsunami.

None of this is to argue that the shock model should be tossed to the curb. Nor is it to suggest that contract language never evolves explosively or at moments punctuated by shocks. It is merely to point out that the process of contractual change can vary depending upon a range of variables—including, but not limited to, the nature of the contract provision at issue, the type of agreement, the identity of the drafter, the geographic location where the contract is prepared, and the centralized or decentralized nature of the drafting community—and that a great deal depends upon context. It is important to resist, in other words, the temptation to view the shock model as a general model of contractual innovation. The shock model provides a compelling explanation for changes to collective action clauses in sovereign debt agreements. It does not, however, do a good job at explaining changes to choice-of-law clauses in other types of contracts over time. This insight, in turn, suggests the need for new models of contractual innovation beyond the shock model.

Conclusion

There are many histories of the doctrine of party autonomy. To date, however, scholars have largely ignored the history of the choice-of-law clause, without which there would be no need for this doctrine. This Article has sought to tell, for the first time, the history of this increasingly common piece of contract boilerplate. It has tracked the prevalence of these clauses over time and has chronicled the interplay between drafters, courts, scholars, and legislatures concerning their enforceability. It has sought to determine how the language in these clauses has evolved and changed over many years. And it has drawn upon the foregoing history to challenge the myth of the rational contract drafter and to call for more differentiated models of contractual innovation. Whether it has achieved each of these goals is for the reader to determine. At a minimum, however, this Article offers answers to a host of basic questions relating to the choice-of-law clause for which the existing literature offers no ready answers.

 

Appendix: Published Case Data Collection

In assembling the data set of 3,104 choice-of-law clauses discussed in Part III of this Article, the following data collection methods were used.

Each research assistant was given a list of search terms keyed to words and phrases often found in choice-of-law clauses and a specific time frame in which to search.[209] They then searched for the terms in the “Cases” database maintained by Lexis Advance for the assigned time frame. When the research assistant found a case containing a choice-of-law clause, he or she entered the name of the case, the year of the contract containing the choice-of-law clause, the type of contract, and the full text of the choice-of-law clause into a spreadsheet. I subsequently reviewed each spreadsheet entry to confirm that the language in question was, in fact, a choice-of-law clause. I also conducted random spot checks to verify the contract year and the type of agreement.

In the overwhelming majority of instances, the clauses identified as part of this search did not play a meaningful role in the court’s analysis. The court simply quoted the text from the clause as part of its background discussion and then moved on to discuss matters unrelated to choice of law.

When the published case did not state the year in which the contract containing the choice-of-law clause was concluded, this particular cell was left blank. After all of the data collection for a particular decade was complete, I calculated the average gap between the year of the case and the year of the contract for those contracts where the date of execution was known. I then assigned the contracts for which the year was unknown an execution date that matched the average for all of the known agreements. With respect to a choice-of-law clause in a contract with an unknown date from a case decided in 1959, for example, the contract was assigned a year of 1951 because the average difference between the date of the case and the date of the known contracts for cases in the 1950s was eight years. With respect to a choice-of-law clause contained in a contract with an unknown date from a case decided in 1934, by contrast, the contract was assigned a year of 1928 because the average difference between the date of the case and the date of the known contracts for cases in the 1930s was six years.

Once the contracts had all been assigned to a decade, I reviewed the language in each of the clauses so that it could be properly coded. In some instances, this task was easy. With respect to the issue of interpretation, for example, I simply counted how many clauses contained variations on the word “interpret.” In other instances, this task was more complicated. With respect to the issue of internal law, for instance, there are many different ways of signaling an intent to select a given state’s internal law. For these clauses, I had to review each clause individually to see if it contained the phrases “internal law,” “without regard to principles of conflict of laws,” “applicable to contracts made and to be performed,” or some other language to similar effect. After all of the clauses had been coded, I then calculated the percentage of clauses for a given decade that contained language responsive to the issue under discussion.

Although I collected data for clauses going back to the nineteenth century, most of the analysis in Part III focuses on clauses from the 1940s onward. There are several reasons for this. First, the clauses dating to the 1880s, 1890s, 1900s, and 1910s were overwhelmingly from a single type of contract—life insurance agreements—and were generally drafted in the same manner. Any general discussion of the clauses from these decades, therefore, would give a misleading impression of broader contract practice because the sample was so dominated by life insurance agreements. Second, the total number of clauses from these early decades was generally small. There were only thirty-one clauses from the 1910s, forty-three clauses from the 1920s, and fifty-three clauses from the 1930s. From that point forward, the number of clauses in the sample increased substantially. There were seventy-nine clauses from the 1940s, 123 clauses from the 1950s, and 236 clauses from the 1960s. Accordingly, the presentation of the data focuses principally on those decades for which there was a larger—and presumably more reliable—body of contracts.

A critic might argue that the sample of clauses here examined is biased because all of them come from published cases. If published cases are not representative of all cases—and they are not—then it may well be that choice-of-law clauses quoted in published cases are not representative of all choice-of-law clauses.[210] This critique is entirely fair. If the goal is to study the changes to the language in choice-of-law clauses over the course of the twentieth century, however, this particular method of data collection is arguably the best of all the bad options. In a perfect world, I would spend a year traveling from company archive to company archive in search of original paper contracts from companies across a wide range of industries in search of choice-of-law clauses. Given the enormous number of contracts negotiated over the past 150 years, however, it is not clear that the sample of clauses gathered through such an endeavor would be any more representative than the ones collected from a review of published cases. Such an inquiry would inevitably omit contracts from companies that failed. It would also omit contracts from companies that do not maintain archives open to researchers.

At the end of the day, the simple truth is that the vast majority of all historical contracts are unavailable to researchers. This means that any attempt to reconstruct past contracting practice must inevitably rely on a sample that may not be representative. This discussion calls to mind the old story of the man looking for his car keys under the streetlamp because it is the only place where he can see to look. While searching through published cases for choice-of-law clauses may be less than ideal, it is the one of the very few places where we can see to look.

  1. * Reef C. Ivey II Distinguished Professor of Law, University of North Carolina at Chapel Hill. Thanks to Ronald Brand, Pam Bookman, Christian Burset, Lauren Coyle, Bill Dodge, Maggie Gardner, Mitu Gulati, Carissa Hessick, Dave Hoffman, Alex Mills, Tanya Monestier, Leigh Osofsky, Steve Sachs, Aaron Simowitz, Kathleen Thomas, Mark Weidemaier, Glenn West, Chris Whytock, and the faculty of the Sandra Day O’Connor School of Law at Arizona State University for comments on an earlier draft of this Article. Thanks to Will Bowers, Kelsey Cullinan, Katie Dixon, Ryan Dovel, Rashaad Hamilton, Richard Hanson, Trevor Kennedy, Ker Medero, Foram Majmudar, M-K McKinney, Marissa Mugan, Jordan Proffitt, Samantha Reeves, Alexandra Russell, Stephen Spivey, Luke Thomas, Andrew Wisniewsky, Edward Woodall, and Christine Xiao for their excellent research assistance. Thanks to Harriet Wu for her help with the figures and tables.
  2. . See Restatement (Second) of Conflict of Laws § 188 (Am. Law Inst. 1971).
  3. . See, e.g., Janet V. Hallahan, The Case of the Missing Decision: When Will Pennsylvania Solve the Mystery of Its “Flexible” Choice-of-Law Analysis?, 69 Temp. L. Rev. 655, 694 (1996).
  4. . See, e.g., Peter Hay, The Use and Determination of Foreign Law in Civil Litigation in the United States, 62 Am. J. Comp. L. 213, 217 (2014); Christopher A. Whytock, Myth of Mess? International Choice of Law in Action, 84 N.Y.U. L. Rev. 719, 732–33 (2009).
  5. . See Mastrobuono v. Shearson Lehman Hutton, Inc., 514 U.S. 52, 59 (1995) (observing that a “choice of law provision, when viewed in isolation, may reasonably be read as merely a substitute for the conflicts-of-laws analysis that otherwise would determine what law to apply to disputes arising out of the contractual relationship”). A recent study found support for the concept of party autonomy in conflict of laws in legal systems all around the world. See Symeon Symeonides, The Scope and Limits of Party Autonomy in International Contracts: A Comparative Analysis, in The Continuing Relevance of Private International Law and New Challenges (Franco Ferrari & Diego P. Fernandez Arroyo eds., 2019).
  6. . Julian Nyarko, We’ll See You in . . . Court! The Lack of Arbitration Clauses in International Commercial Contracts, 58 Int’l Rev. L. & Econ. 6, 11 (2018).
  7. . Interestingly, the history of the forum selection clause has garnered more academic attention than the history of the choice-of-law clause notwithstanding the fact that the latter was (and is) more common than the former. See David Marcus, The Perils of Contract Procedure: A Revised History of Forum Selection Clauses in the Federal Courts, 82 Tul. L. Rev. 973, 996–1014 (2008) (discussing the history of the forum selection clause).
  8. . See, e.g., Alex Mills, Party Autonomy in Private International Law 44–64 (2018); Symeon Symeonides, Codifying Choice of Law Around the World: An International Comparative Analysis 110–15 (2014); Matthias Lehman, Liberating the Individual from Battles Between States: Justifying Party Autonomy in Conflict of Laws, 41 Vand. J. Transnat’l L. 381, 385–90 (2008); Gisela Ruhl, Party Autonomy in the Private International Law of Contracts: Transnational Convergence and Economic Efficiency, in Conflict of Laws in a Globalized World 4–8 (Eckart Gottschalk ed., 2007).
  9. . Similarly, there is no need for any doctrine of party autonomy with respect to the chosen forum absent a forum selection clause or an arbitration clause.
  10. . As an empirical matter, it is a near-impossible task to provide a comprehensive assessment of contract drafting practice as it existed more than a century ago. Each of the sources listed in this paragraph, therefore, represents a second-best attempt to determine what such practice looked like. Collectively, they constitute the most comprehensive data set possible that can help us make inferences about party behavior, notwithstanding the limitations inherent in their use.
  11. . Joseph Story, Commentaries on the Conflict of Laws § 233 (1834).
  12. . Id.
  13. . Id.
  14. . See id. § 278.
  15. . Robinson v. Bland, Eng. Rep. 717 (K.B. 1760) (emphasis added).
  16. . Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 48 (1825) (Marshall, C.J.) (emphasis added); see also Morris J. Levin, Party Autonomy: Choice-of-Law Clauses in Commercial Contracts, 46 Geo. L.J. 260, 269 (1957).
  17. . See infra note 19 and accompanying text (discussing the first express choice-of-law clauses); see also Carnegie v. Morrison, 43 Mass. 381, 401 (1841) (“There is no reference, tacit or express, in this instrument, to the laws of England, which can raise a presumption, that the parties looked to them as furnishing the rule of law, which should govern this contract.” (emphasis added)).
  18. . There are a small number of cases in which the court relied upon writings in non-usury contracts to ascertain the intent of the parties. See, e.g., Strawbridge v. Robinson, 10 Ill. 470, 472–73 (1849) (concluding that the parties had expressly selected the law of Illinois by writing “East Fork the 4th Feb’y 1845” on the face of their agreement).
  19. . Butters v. Olds, 11 Iowa 1, 2–3 (1860); see also Scott v. Perlee, 39 Ohio St. 63, 66–67 (Ohio 1883) (citations omitted) (“[I]t is undoubtedly the law of this state, and indeed it is now well established almost universally, that where a contract is entered into in one state, to be performed in another, between citizens of each, and the rate of interest is different in the two, the parties may, in good faith, stipulate for the rate of either, and thus expressly determine with reference to the law of which place that part of the contract shall be decided.”); Peck v. Mayo, Follett & Co., 14 Vt. 33, 38 (1842) (“If a contract be entered into in one place to be performed in another, and the rate of interest differ in the two countries, the parties may stipulate for the rate of interest of either country, and thus, by their own express contract, determine with reference to the law of which country that incident of the contract shall be decided.”).
  20. . Kirtland v. Hotchkiss, 42 Conn. 426, 444 (1875). This is not to suggest that the express choice-of-law clause sprang—like Athena—fully formed from the head of Zeus in 1869. There are clauses in English contracts from the 1760s directing arbitrators to apply the “Rules and Customs” of a particular place. See Christian Burset, Arbitrating the England Problem: Litigation, Private Ordering, and the Rise of the Modern Economy, 36 Ohio St. J. on Disp. Resol. (forthcoming 2020), https://ssrn.com/abstract=3561136 [https://perma.cc/B8DT-V5HP]. And choice-of-law clauses can be found in foreign maritime contracts dating to 1800 that select the law of a particular place. See Aertsen v. The Aurora, 1 F. Cas. 206, 207 (D.S.C. 1800) (Case No. 95) (referencing document selecting the “marine law of Hamburgh” to regulate “the conduct of officers and seamen aboard vessels belonging to that place”). It is merely to state that the express choice-of-law clause in its modern incarnation dates to roughly 1869.
  21. . Kirtland, 42 Conn. at 444; see also Warner v. Warner, 235 Ill. 448, 456 (1908) (referencing a choice-of-law clause in an 1874 prenuptial agreement).
  22. . Union Cent. L. Ins. Co. v. Pollard, 94 Va. 146, 151 (1896).
  23. . New York Life Ins. Co. v. Cravens, 178 U.S. 389, 395 (1900).
  24. . Mutual Ben. Life. Ins. Co. v. Robison, 54 F. 580, 582 (N.D. Iowa 1893).
  25. . The Oranmore, 24 F. 922, 923 (D. Md. 1885). The judge in this case specifically noted that the express choice-of-law clause was a provision “which I have not met with before.” Id.
  26. . The Kensington, 183 U.S. 263, 269 (1902).
  27. . Id.
  28. . New Eng. Mortg. Sec. Co. v. Vader, 28 F. 265, 275 (C.C.D. Or. 1886); see also Farrior v. New Eng. Mortg. Sec. Co., 88 Ala. 275, 277 (1889) (stating that “the notes herein described, and this mortgage, shall be governed and construed by and under the laws of the State of Alabama, where the same is made”).
  29. . Padley v. Neill, 134 Mo. 364, 369–70 (1896).
  30. . Smith v. Parsons, 55 Minn. 520, 526 (1893).
  31. . In 1882, the Supreme Court reiterated its support for the general principle of party autonomy in a case that did not contain an express choice-of-law clause. See Pritchard v. Norton, 106 U.S. 124, 137 (1882) (“It is always to be remembered that in obligations it is the will of the contracting parties, and not the law, which fixes the place of fulfilment—whether that place be fixed by express words or by tacit implication—as the place to the jurisdiction of which the contracting parties elected to submit themselves.”).
  32. . See Lanier v. Union Mortg. Banking & Tr. Co., 40 S.W. 466, 469 (Ark. 1897); Wilson v. Lewiston Mill Co., 150 N.Y. 314, 323 (1896); Union Cent. Life Ins. v. Pollard, 26 S.E. 421, 422 (Va. 1896); Smith v. Parsons, 57 N.W. 311, 313 (Minn. 1893); New Eng. Mortg. Sec. Co. v. McLaughlin, 13 S.E. 81, 83 (Ga. 1891).
  33. . Sometimes the problems were more prosaic. In 1900, for example, the Supreme Court observed that “a contract by an insurance company of New York executed elsewhere may by its terms incorporate the law of New York, and make its provisions controlling upon both the insured and the insurer.” Mutual Life Ins. v. Cohen, 179 U.S. 262, 267 (1900). On the facts of the case before it, however, the Court held that because the choice-of-law clause had been written into the application for a life insurance policy rather than the policy itself, the parties had not, in fact, chosen the law of New York to govern their agreement. Id. at 270.
  34. . Fletcher v. N.Y. Life Ins. Co., 13 F. 526, 528 (1882) (emphasis added); see also Francis Wharton, A Treatise on the Conflict of Laws 1194 (1905) (“[I]f a borrower residing in one state and a lender in another enter into a contract in the latter state, making it payable in the former, and fix a rate of interest greater than allowed by the law of either state, but expressly stipulate that the contract shall be governed by the law of a third state where such rate is legal, it is apparent, at least in the absence of additional circumstances showing that some of the elements of the contract or significant circumstances of the transaction have their situs in that state, that the parties acted in bad faith and for the purpose of evading the usury law of one or both of the other states.”).
  35. . The Energia, 56 F. 124, 127 (S.D.N.Y. 1893) (citations omitted).
  36. . Lindenberger Cold Storage & Canning Co. v. J. Lindenberger, Inc., 235 F. 542, 547 (W.D. Wash. 1916) (stipulating that the agreement was “to be construed and take effect as a contract made in England and in accordance with the law of England”).
  37. . Weil v. Chi. Pneumatic Tool Co., 212 S.W. 313, 314 (Ark. 1919) (agreement stipulating that “[it] shall be interpreted and construed according to the Laws of the State of Illinois”).
  38. . Chalmers & Williams v. Surprise, 123 N.E. 841, 842 (Ind. App. 1919) (stating that “the contract shall be deemed consummated at Chicago, Illinois”).
  39. . Carl Hagenback & Great Wallace Show Co. v. Randall, 126 N.E. 501, 502 (Ind. 1920).
  40. . Id.; see also Pollak v. Danbury Mfg. Co., 131 A. 426, 428 (Conn. 1925) (discussing choice-of-law clause written into the employment contract of a general manager at a factory); Laure v. Singer, 100 N.J.L. 98, 100 (N.J. 1924) (discussing choice-of-law clause written into the employment contract of a theatrical performer).
  41. . See Shafer v. Metro-Goldwyn-Mayer Distrib. Corp., 172 N.E. 689, 691 (Ohio 1929); see also Erie, Bernhardt, and Section 2 of the United States Arbitration Act: A Farrago of Rights, Remedies, and a Right to a Remedy, 69 Yale L.J. 847, 854 n.41 (1960) (discussing history of the New York arbitration law that provided the basis for the Federal Arbitration Act).
  42. . See, e.g., Binderup v. Pathe Exch., Inc., 263 U.S. 291, 302 (1923).
  43. . See Warner v. Warner, 85 N.E. 630, 633–34 (Ill. 1908) (referencing a choice-of-law clause in an 1874 prenuptial agreement).
  44. . Swan v. Swan’s Ex’r, 117 S.E. 858, 859 (Va. 1923).
  45. . Id.
  46. . Boggs v. Boggs, 114 A. 474, 475 (Md. 1921).
  47. . Id.
  48. . Id.
  49. . In the contract form books from this era, for example, less than 2 percent of the agreements contain choice-of-law clauses. See infra Section I.F (discussing form books from the 1910s).
  50. . Although the first life insurance contract to contain a choice-of-law clause dates to 1880, and although most life insurance contracts contained these clauses in 1902, many life insurance contracts issued in the years after 1905 lacked choice-of-law clauses. Compare Allen J. Flitcraft, Life Insurance Manual (15th ed. 1902) (collecting forty-eight life insurance agreements and applications, of which 66 percent contained choice-of-law clauses) with Allen J. Flitcraft, Life Insurance Manual (23rd ed. 1910) (collecting fifty-nine life insurance agreements or applications, of which 7 percent contained choice-of-law clauses). This shift in contract practice is attributable to the enactment of state laws between 1906 and 1909 mandating that insurance companies use standard policy forms approved by the state. See R. Carlyle Buley, I The American Life Convention, 1906–1952: A Study in the History of Life Insurance 274, 280, 310 (1953). These standard policy forms generally lacked choice-of-law clauses. The state legislation thus helps to explain why some of the best known conflicts decisions rendered by the U.S. Supreme Court in the twentieth century involved insurance contracts that lacked a choice-of-law clause notwithstanding the fact that these clauses could be commonly found in policies issued earlier in the century. See, e.g., Allstate Ins. Co. v. Hague, 449 U.S. 302 (1981); Clay v. Sun Ins. Office, 377 U.S. 179 (1964).
  51. . Kirtland v. Hotchkiss, 42 Conn. 426, 444 (1875).
  52. . Gile v. Interstate Motor Car Co., 145 N.W. 732, 733 (N.D. 1914).
  53. . U.S. Mortg. Co. v. Sperry, 138 U.S. 313, 338 (1891).
  54. . Lindenberger Cold Storage & Canning Co. v. J. Lindenberger, Inc., 235 F. 542, 547 (W.D. Wash. 1916).
  55. . Warner v. Warner, 85 N.E. 630, 633–34 (Ill. 1908).
  56. . Binderup v. Pathe Exch. Inc., 263 U.S. 291, 302 (1923).
  57. . Equitable Life Assurance Soc. v. Clements, 140 U.S. 226 (1891).
  58. . Chalmers & Williams v. Surprise, 123 N.E. 841, 842 (Ind. App. 1919).
  59. . The Oranmore, 24 F. 922, 923 (D. Md. 1885).
  60. . Weisberg v. Hunt, 131 N.E. 471, 472 (Mass. 1921).
  61. . The Kensington, 183 U.S. 263, 269 (1902).
  62. . Carl Hagenbeck & Great Wallace Show Co. v. Randall, 126 N.E. 501, 502 (Ind. App. 1920).
  63. . Comm’r v. Bateman, 127 F.2d 266, 268 (1st Cir. 1942).
  64. . State v. Hall, 114 S.E. 250, 251 (W. Va. 1922).
  65. . Thoms v. Thoms, 222 Ill. App. 618, 622 (1921).
  66. . R.C.A. Photophone, Inc. v. Sinnott, 30 P.2d 761, 762 (Or. 1934).
  67. . Samuel Williston, Joseph Henry Beale: A Biographical Sketch, 56 Harv. L. Rev. 685 (1943); see also Symeon Symeonides, The First Conflicts Restatement Through the Eyes of the Old: As Bad as Its Reputation?, 32 S. Ill. U. L.J. 39, 41–45 (2007).
  68. . Joseph H. Beale, What Law Governs the Validity of a Contract, 23 Harv. L. Rev. 1 (1909). Beale argued that “the doctrine itself is one that is quite foreign to common-law notions.” Id. at 7. He argued further that the idea that contracting parties should be allowed to choose a law to govern their contracts “is not a natural notion in a law based like ours on the complete jurisdiction of the territorial sovereign.” Id.
  69. . Joseph H. Beale, II The Conflict of Laws 1095 (1935).
  70. . Id. at 1096.
  71. . Id. at 1105–09 (identifying several U.S. Supreme Court decisions in which the Court resolved a choice-of-law issue without specifically stating that the intent of the parties should be considered).
  72. . Id. at 1085–86.
  73. . Joseph Beale, A Treatise on the Conflict of Laws 107 (1916) (“A right having been created by the appropriate law, the recognition of its existence should follow everywhere.”).
  74. . See Ala. Great S. R.R. v. Carroll, 97 Ala. 126 (1892).
  75. . Restatement (First) of Conflict of Laws § 1 (Am. Law Inst. 1934).
  76. . Beale, supra note 68, at 1079–80.
  77. . Hay et al., Conflict of Laws 21 (5th ed. 2010).
  78. . See Symeonides, supra note 66, at 66–74 (discussing the drafting process within the American Law Institute that led to the production of the First Restatement and explaining how Beale managed to push through a draft that was inconsistent with the case law in many respects).
  79. . The one concession that Beale was willing to make involved contract interpretation. Even he acknowledged that “the parties’ intention governs in this matter of the meaning of the words used.” Beale, supra note 68, at 1201. He believed, however, that interpretation isa wholly different point from that involved in allowing them to choose the law to govern the obligation. Here there is no question of validity; the sole question is that of fixing the meaning of language and there is no more difficulty in giving effect to a Texas use of a term in a Missouri contract than in giving effect to French or German instead of English as a means of expression.Id. These observations were made in his treatise. They were not expressed in the First Restatement.
  80. . Id. at 1174.
  81. . Id.
  82. . Arthur Nussbaum, Conflict Theories of Contracts: Cases Versus Restatement, 51 Yale L.J. 893 (1942).
  83. . Id. at 895. “Private international law” is the phrase that non-U.S. legal scholars use to refer to the body of doctrine that is known as conflict of laws in the United States. See Zicherman v. Korean Air Lines Co., 516 U.S. 217, 227 (1996).
  84. . Nussbaum, supra note 81.
  85. . Id. at 922.
  86. . In 1928, Judge Finch warned Beale that the judges would not blindly follow Beale’s position on this issue even if he wrote it into the First Restatement. See 6 American Law Institute Proceedings 458 (1927–28) (“[Y]‌ou will never be able to hold your courts to that sort of a rule. You can lay it down, but human nature is not so constituted that you can make a court adopt a general rule which will do injustice in a majority of the cases coming with it.”) (quoted in Symeonides, supra note 66, at 113); see also Mills, supra note 7, at 61 (“The rejection of party autonomy, even by such influential figures, did not however halt its rise in practice.”).
  87. . Boseman v. Conn. Gen. Life Ins., 301 U.S. 196, 202 (1937). At about this same time, the U.S. Supreme Court issued several other decisions making clear that the U.S. Constitution did not mandate a reliance on the vested rights theory in choice of law. See James Y. Stern, Choice of Law, the Constitution, and Lochner, 94 Va. L. Rev. 1509, 1525 (2008).
  88. . Hal Roach Studios, Inc. v. Film Classics, Inc., 156 F.2d 596, 598 (2d Cir. 1946) (internal citations and quotation marks omitted). This decision had the effect of overturning two prior Second Circuit decisions authored by Judge Learned Hand criticizing the practice of party autonomy. See E. Gerli & Co. v. Cunard S.S. Co., 48 F.2d 115 (2d Cir. 1931); Louis-Dreyfus v. Paterson S.S., Ltd., 43 F.2d 824 (2d Cir. 1930).
  89. . See also Lauritzen v. Larsen, 345 U.S. 571, 588–89 (1953) (“[T]his contract was explicit that the Danish law and the contract with the Danish union were to control. Except as forbidden by some public policy, the tendency of the law is to apply in contract matters the law which the parties intended to apply. We are aware of no public policy that would prevent the parties to this contract, which contemplates performance in a multitude of territorial jurisdictions and on the high seas, from so settling upon the law of the flag-state as their governing code.” (emphasis added)).
  90. . U.C.C. § 1-105 (Am. Law Inst. 1952).
  91. . This is not the first article to look to form books to answer other questions about historical contract practice. See Lee C. Buchheit & G. Mitu Gulati, Sovereign Bonds and the Collective Will, 51 Emory L.J. 1317, 1325 (2004) (relying on English form books to determine the first use of the collective action clause).
  92. . Adam Freedman, The Party of the First Part: The Curious World of Legalese 25 (2007) (“Almost as soon as Gutenberg’s first Bible rolled off the press (1455), English lawyers were putting together formbooks, that is, collections of sample contracts, pleadings, and other documents that had already passed muster with some court or another. Provided that one copied the form verbatim, no sporting judge could object.”); M.H. Hoeflich, Law Blanks & Form Books: A Chapter in the Early History of Document Production, 11 Green Bag 2d 189, 191 (2008) (“The time-saving aspect of the use of forms was particularly important to American lawyers for much of the nineteenth century because they tended to bill by the transaction, rather than by the hour or increment thereof.”).
  93. . See William E. Foster & Andrew L. Lawson, When to Praise the Machine: The Promise and Perils of Automated Transactional Drafting, 69 S.C. L. Rev. 597, 613–23 (2018) (discussing the changing drafting practices of corporate lawyers over time and the changing role of form books in the computer age).
  94. . Graybill v. Graybill, 14 Pa. D. & C. 382, 384 (C.P. 1930).
  95. . See Nyarko, supra note 5, at 11.
  96. . J.R. Sayler, Sayler’s American Form Book (1913) (250 contracts; 0 clauses); Francis B. Tiffany, Legal and Business Forms (1915) (221 contracts; 2 clauses).
  97. . Clarence F. Birdseye, Encyclopaedia of General Business and Legal Forms (1924) (636 contracts; 11 clauses); Leslie M. O’Connor, Pocket Manual of Standard Legal and Business Forms (1921) (95 contracts; 1 clause); William Herbert Page, The Law of Contracts (1922) (170 contracts; 1 clause).
  98. . Edmund O. Belsheim, I Modern Legal Forms 413–554 (1938); Edmund O. Belsheim, IV Modern Legal Forms 1–438 (1938) (263 total contracts; 6 total clauses).
  99. . Saul Gordon, Gordon’s Modern Annotated Forms of Agreement (1946) (431 contracts; 11 clauses).
  100. . Leonard A. Jones, Legal Forms Annotated: Contractual Business and Conveyancing Forms (1946) (678 contracts; 3 clauses).
  101. . Even if the publication of the First Restatement had generated widespread concern about the enforceability of choice-of-law clauses, there was arguably no real harm to leaving existing clauses in agreements. After all, if a court refused to enforce them, then the parties were no worse off than before.
  102. . I Am. Jur. Legal Forms Annotated (1953) (132 contracts; 1 clause).
  103. . William J. Casey, Forms of Business Agreements with Tax Ideas (1961).
  104. . David W. Carroll, Forms for Commercial Transactions Under the Uniform Commercial Code (1963).
  105. . While it is also possible that the tremendous expansion of commerce after the Second World War led to more contracts between commercial actors in different states and that these dealings with out-of-state parties, in turn, may have led a growing number of parties to write choice-of-law clauses into their agreements, it is impossible to know for sure the extent to which the expansion of commerce during these years contributed to the more extensive use of choice-of-law clauses.
  106. . See Robert J. Nordstrom & Dale B. Ramerman, The Uniform Commercial Code and the Choice of Law, 1969 Duke L.J. 624, 636 n.45 (1969).
  107. . It is unlikely, however, that this increase was driven by a new belief that such clauses were enforceable. As discussed above, the caselaw post-1934 had broken decisively in favor of enforcement such that it would have been unreasonable for contract drafters to consciously omit such clauses from their contracts. It is far more likely that the furor relating to the enactment of the UCC—and the attention given to section 1-105—enhanced the perceived salience of choice-of-law clauses. More lawyers were thinking about choice-of-law clauses because they were made aware of section 1-105. This enhanced salience, in turn, may have prompted attorneys to revise their standard forms to incorporate choice-of-law clauses.
  108. . Restatement (Second) of Conflict of Laws § 187(1) (Am. Law Inst. 1971). A default rule is one that the parties can contract around in their agreement. The parties may, for example, select the law of any state to govern issues relating to contract interpretation because they could just as easily rewrite their agreement to resolve the interpretive issue themselves.
  109. . Id. § 187(2). A mandatory rule is one that the parties cannot contract around in their agreement. The parties generally may not, for example, select the law of just any state to govern issues relating to securities laws because most states require that their securities laws be applied to certain transactions.
  110. . Id.
  111. . Id. cmt. f.
  112. . Id.
  113. . Id. cmt. g; see also supra notes 33–34 and accompanying text.
  114. . In the years since the Second Restatement was published, section 187 has proven to be popular among U.S. courts. Hay et al., supra note 76, at 75 (describing section 187 as “one of the Restatement’s most successful and popular provisions”). Even when a particular jurisdiction has declined to follow the Second Restatement with respect to other issues, its courts will frequently look to section 187 to determine the enforceability of a choice-of-law clause. Id. at 1088 (observing that section 187 “is followed by more American courts than any other provision in the Restatement (Second), including some courts that otherwise follow the traditional theory”) (internal quotation marks and citations omitted).
  115. . The statute provided:if a policyholder had made at least two timely payments of premiums, defaulted on a subsequent premium, and died within sixty days of the default, the beneficiary [could] recover the accrued value of the policy “as if there had been no default in the payment of premium, anything in the policy to the contrary notwithstanding.”Clyde Spillenger, Risk Regulation, Extraterritoriality, and Domicile: The Constitutionalization of American Choice of Law, 1850–1940, 62 UCLA L. Rev. 1240, 1298 (2015) (emphasis added).
  116. . See N.Y. Life Ins. v. Cravens, 178 U.S. 389, 401 (1900). Even today, a few states have laws on the books that require state courts to apply state law to all contracts insuring lives, property, or interests within the state. See, e.g., N.C. Gen. Stat. Ann. § 58-3-1 (West 2019); S.C. Code Ann. § 38-61-10 (West 2019); Va. Code. Ann. § 38.2-313 (West 2019).
  117. . See Unif. Consumer Credit Code § 1.201(8). This Code has been adopted by eleven states.
  118. . See George F. Carpinello, Testing the Limits of Choice of Law Clauses: Franchise Contracts as a Case Study, 74 Marq. L. Rev. 57, 71 n.68 (1990) (listing states that have enacted franchise statutes containing anti-waiver provisions).
  119. . See, e.g., Cal. Lab. Code § 925(a) (West 2019) (“An employer shall not require an employee who primarily resides and works in California, as a condition of employment, to agree to a provision that would . . . [d]eprive the employee of the substantive protection of California law with respect to a controversy arising in California.”); Kan. Stat. Ann. § 50-625(a) (West 2019) (“Except as otherwise provided in this act, a consumer may not waive or agree to forego rights or benefits under this act.”); see also La. Stat. Ann. § 23:921(A)(2) (West 2019) (“The provisions of every employment contract or agreement . . . by which any foreign or domestic employer or any other person or entity includes a . . . choice of law clause in an employee’s contract of employment . . . shall be null and void except where the . . . choice of law clause is expressly, knowingly, and voluntarily agreed to and ratified by the employee after the occurrence of the incident which is the subject of the civil or administrative action.”).
  120. . See generally Hay et al., supra note 76, at 1108–26 (discussing public policy limitations across a range of different types of agreements).
  121. . See Cyra Akila Choudhury, Shari’ah Law as National Security Threat?, 46 Akron L. Rev. 49, 52–65 (2013).
  122. . See, e.g., Ark. Code Ann. § 1-1-103 (2019). In some cases, the legislation distinguished between commercial contracts and noncommercial contracts. See N.C. Gen. Stat. § 1-87.14 (2019) (“A court . . . shall not apply a foreign law in any legal proceeding involving . . . a claim for absolute divorce, divorce from bed and board, child custody, child support, alimony, or equitable distribution if doing so would violate a fundamental constitutional right of one or more natural persons who are parties to the proceeding.”). In other cases, it did not. See Kan. Stat. Ann. § 60-5104 (2019). In a number of instances, the legislation provided that the prohibition on applying foreign law would not apply to artificial entities such as corporations. See, e.g., Ariz. Rev. Stat. Ann. § 12-3102 (2019); La. Stat. Ann. §9:6001(G) (2019); Kan. Stat. Ann. § 60-5108 (2019).
  123. . Ex parte Cont’l Motors, Inc., 270 So. 3d 1148, 1151 (Ala. 2018).
  124. . N.Y. Gen. Oblig. L. § 5-1401(1) (McKinney 1984). The statute stated that it would not apply to any contract for labor or personal services, any contract for personal, family, or household services, or contracts implicating issues covered by certain sections of the Uniform Commercial Code. Id. § 5-1401(2).
  125. . See Theodore Eisenberg & Geoffrey P. Miller, The Market for Contracts, 30 Cardozo L. Rev. 2073, 2091 (2009).
  126. . See, e.g., Cal. Civ. Code § 1646.5 (West 2019); Del. Code Ann. Tit. 6, § 2708 (West 2019); Fla. Stat. Ann. § 685.101 (West 2019); 735 Ill. Comp. Stat. Ann. 105‌/5-5 (West 2019); Ohio Rev. Code Ann. § 2307.39 (West 2019).
  127. . N.C. Gen. Stat. Ann. § 1G-3(a)(1) (West 2019). The statute defines a business contract as “a contract or undertaking, contingent or otherwise, entered into primarily for business or commercial purposes. The term does not include a consumer contract or employment contract.” Id. § 1G-2.
  128. . Id. § 1G-3(a)(2).
  129. . .Sarath Sanga, Choice of Law: An Empirical Analysis, 11 J. Empirical Legal Stud. 894, 90203 (2014).
  130. . Nyarko, supra note 5, at 10–11.
  131. . John F. Coyle & Christopher R. Drahozal, An Empirical Study of Dispute Resolution Clauses in International Supply Contracts, 52 Vand. J. Transnat’l L. 323 (2019); see also Gilles Cuniberti, The International Market for Contracts: The Most Attractive Contract Laws, 34 Nw. J. Int’l L. & Bus. 455, 459, 469 (2014) (“A study of contracts in cases that came before the International Court of Arbitration of the International Chamber of Commerce between 2007 and 2012 found that 83% contained choice-of-law clauses.”).
  132. . Kirtland v. Hotchkiss, 42 Conn. 426, 444 (1875). In order to appreciate how little has changed in some respects since the 1860s, compare this clause to the following clause from 2015: “[This Agreement] shall be deemed to be a contract made under the laws of the State of New York, and for all purposes shall be governed by and construed in accordance with the laws of said State.” Carnival PLC and U.S. Bank Nat’l Ass’n, Indenture (Form 305B2) (Nov. 6, 2015); see also Fairmount Santrol Holdings Inc. and The Bank of N.Y. Mellon Tr. Co., Indenture (Form 305B2) (Dec. 1, 2016) (“This Indenture and the Securities will be deemed to be a contract made under the laws of the State of New York, and for all purposes will be construed in accordance with the laws of said State without giving effect to principles of conflicts of laws of such State.”).
  133. . We also collected 243 clauses that selected the law of a foreign jurisdiction. Since this Article is focused exclusively on U.S. contract practice, and since most choice-of-law clauses selecting a foreign jurisdiction were drafted by foreign parties with an eye to foreign law, these foreign clauses were excluded from the data set.
  134. . Bailey v. Jerome, 129 S.C. 387 (1924) (emphasis added); see also Farmers’ & Breeders’ Mut. Reserve Fund Live Stock Ins. v. Olson, 22 Pa. D. 437 (C.P. 1913) (“This policy of insurance is a contract made and to be performed in Williamsport, Lycoming County, Pennsylvania.”).
  135. . Gleason v. Nw. Mut. Life Ins., 203 N.Y. 507, 510 (1911).
  136. . Fountain v. Sec. Mut. Life Ins., 93 S.E. 118, 119 (Ga. 1917).
  137. . Hay et al., supra note 76, at 19–24.
  138. . Id. at 65–75.
  139. . Id. at 67.
  140. . Id. at 72.
  141. . Id. at 58–63.
  142. . Restatement (Second) of Conflict of Laws § 188(1) (Am. Law Inst. 1971) (emphasis added).
  143. . It is important to note that clauses that contain language keyed to vested rights have always comprised a relatively small percentage of all choice-of-law clauses. It is much more common for such clauses to utilize the words “interpret,” “construe,” or “govern” than to explicitly reference a specific place, as discussed in Sections II.2 and II.3.
  144. . William F. Elliott, Commentaries on the Law of Contracts 774 (1913).
  145. . Id.
  146. . Dwight Arven Jones, Construction or Interpretation of Commercial and Trade Contracts 3 (1886) (emphasis added).
  147. . Elliott, supra note 143, at 774.
  148. . Louis L. Hammond, General Principles of the Law of Contract 775 (1902).
  149. . Joseph Story, Commentaries on the Conflict of Laws 232 (1834).
  150. . Wharton, supra note 33, at 884.
  151. . Beale, supra note 68, at 1201–05.
  152. . The drafters of the Second Restatement took the position that the courts must always apply the law of the forum to interpret a contract. See Restatement (Second) of Conflict of Laws § 204 cmt. a (Am. Law Inst. 1971). If the meaning of the contract is ambiguous, however, then the courts should seek to determine which law should be used in order to construe the agreement. Where the contract contains a choice-of-law clause, the court should then apply the law chosen by the parties to construe the contract. Id. § 204 cmt. b. Section 204 has been cited twelve times since it was published. In none of these cases did the courts distinguish between the act of “interpreting” a contract and the act of “construing” it.
  153. . See infra note 153 (collecting sources); cf. H. G. Craig & Co. v. Uncas Paper Bd. Co., 133 A. 673, 675 (Conn. 1926) (“The appellant . . . contends that the word ‘made’ should be construed as ‘governed.’ We are unable to adopt the latter contention; the only meaning of which ‘made’ as here used is fairly susceptible is as relating to the place of signing and delivery and as thus fixing the locus contractus as being in Pennsylvania.”).
  154. . See Boat Town U.S.A., Inc. v. Mercury Marine Div. of Brunswick Corp., 364 So. 2d 15, 17 (Fla. Dist. Ct. App. 1978) (“In the instant case, there is no assertion, nor could any be substantiated, that ambiguities exist in the terms of the contract. Thus, the interpretation clause of the contract has no effect and does not provide an explicit choice of Wisconsin law to govern the conduct of the parties.”); see also Arnone v. Aetna Life Ins., 860 F.3d 97, 107–08 (2d Cir. 2017) (“In the context presented here, that provision [stating that the agreement will be ‘construed’ in accordance with Connecticut law] is insufficient to bind this court to apply the full breadth of Connecticut law, to the exclusion of another jurisdiction’s law, in fields other than the interpretation of the language in this contract.”); Am.’s Favorite Chicken Co. v. Cajun Enters., 130 F.3d 180, 182 (5th Cir. 1997) (“Since the . . . claims do not implicate the interpretation or construction of the franchise agreements, they are not governed by the narrow choice of law clause present here.”); Dollar Sys., Inc. v. Avcar Leasing Sys., Inc., 890 F.2d 165, 171 (9th Cir. 1989) (observing that the plaintiff’s “compliance with franchise law did not depend on the construction of the license agreement” and that the choice-of-law clause was therefore inapplicable); Procter v. Mavis, 125 P.3d 801, 803 (Or. 2005) (“The choice-of-laws provision in the parties’ premarital agreement is just that: an agreement that California law will govern the construction of the agreement. The provision does not relate to the law applicable to the division of property on dissolution.” (emphasis in original)). But see Hammel v. Ziegler Fin. Corp., 334 N.W.2d 913, 916 (Wis. Ct. App. 1983) (“We . . . can conceive of few instances where it would be reasonable to look to the law of a specific state to define contractual terms but to the law of a second jurisdiction to ascertain the legal effect of the agreement. Such a maneuver would be unreasonable because the meaning associated with a term by one jurisdiction might not mesh with the statutory and common-law scheme of another.”); see also John F. Coyle, The Canons of Construction for Choice-of-Law Clauses, 92 Wash. L. Rev. 631, 656–61 (2017) [hereinafter Coyle, Canons of Construction].
  155. . Boatland, Inc. v. Brunswick Corp., 558 F.2d 818 (6th Cir. 1977); C.A. May Marine Supply Co. v. Brunswick Co., 557 F.2d 1163 (5th Cir. 1977).
  156. . Knieriemen v. Bache Halsey Stuart Shields, Inc., 427 N.Y.S.2d 10, 12 (N.Y. App. Div. 1980).
  157. . Id.
  158. . Id. This outcome was consistent with those reached by other courts asked to decide whether a choice-of-law clause encompassed the tort law of the chosen jurisdiction in the late 1970s. See Travis v. Harris Corp., 565 F.2d 443, 446 (7th Cir. 1977) (“The parties and the district court agreed that Ohio law should be applied because the 1964 contract so provided. Though the contract may be interpreted under Ohio law, the legal effect of that agreement, and questions of traditional tort law unrelated to the contract, are to be determined in accord with the laws of Indiana, the situs of the injury and domicile of Travis.”); Indus. Consultants, Inc. v. HS Equities, Inc., 1980 U.S. Dist. LEXIS 10961, at *8 (S.D.N.Y. Apr. 14, 1980) (concluding choice-of-law clause stating that the agreement was to be “construed” in accordance with New York law did not cover action for misrepresentation); Barron v. Kane & Roach, Inc., 398 N.E.2d 244, 246 (Ill. 1979) (“The question before us . . . is unrelated to a construction of the provisions of the contract. That question of traditional tort law is to be determined in accordance with the law of Illinois, the situs of the injury and the domicile of the injured party.”).
  159. . See Peter A. Alces, Guerilla Terms, 56 Emory L.J. 1511, 1513 (2007) (observing that “it is not in rational form drafters’ interest” to bring [one-sided contract provisions] to the attention of less sophisticated consumers”); Robert Prentice, Contract-Based Defenses in Securities Fraud Litigation: A Behavioral Analysis, 2003 U. Ill. L. Rev. 337, 386 (2003) (“Farnsworth . . . has noted that in his own experience in legal practice, ‘no one in any of the corporations or in the law firm ever suggested that the forms should be drafted other than as one-sidedly in the interests of the corporate client as possible.’” (quoting E. Allan Farnsworth, On Trying to Keep One’s Promises: The Duty of Best Efforts in Contract Law, 46 U. Pitt. L. Rev. 1, 44 (1984))).
  160. . Salovaara v. Jackson Nat’l Life Ins., 66 F. Supp 2d 593, 596 (D.N.J. 1999) (“related . . . to”); Anderson v. First Commodity Corp., 618 F. Supp. 262, 264 (W.D. Wis. 1985) (“in connection with”).
  161. . Rooney v. Biomet, Inc., 63 F. Supp. 2d 126, 127 (D. Mass. 1999) (“relationship”); Am. Standard Leasing Co. v. Plant Specialties, Inc., 427 So. 2d 555, 556 (La. Ct. App. 1983) (“tort”).
  162. . See, e.g., Krock v. Lipsay, 97 F.3d 640, 645 (2d Cir. 1996) (“Under New York law, in order for a choice-of-law provision to apply to claims for tort arising incident to the contract, the express language of the provision must be ‘sufficiently broad’ as to encompass the entire relationship between the contracting parties.”).
  163. . It appears that the first two decades of the twenty-first century have witnessed a shift in this regard. A recent review of 351 choice-of-law clauses written into bond indentures filed with the SEC in 2016 found that approximately 12 percent contained language intended to give the clause a broader scope. John F. Coyle, Choice-of-Law Clauses in U.S. Bond Indentures, 13 Cap. Mkts. L.J. 152 (2018) [hereinafter Coyle, U.S. Bond Indentures]. A separate review of 159 international supply agreements filed with the SEC between 2011 and 2015 found that approximately 22 percent contained similar language. John F. Coyle & W. Mark C. Weidemaier, Interpreting Contracts Without Context, 67 Am. U. L. Rev. 1673 (2018).
  164. . Carl Hagenbeck & Great Wallace Show Co. v. Randall, 126 N.E. 501, 502 (Ind. App. 1920) (emphasis added).
  165. . See Michael Gruson, Governing Law Clauses Excluding Principles of Conflict of Laws, 37 Int’l L. 1023, 1025 (2003).
  166. . Duskin v. Pa.-Cent. Airlines Corp. 167 F.2d 727, 732 (6th Cir.), cert. denied, 335 U.S. 829 (1948).
  167. . See Siegelman v. Cunard White Star Ltd., 221 F.2d 189, 194–96 (2d Cir. 1955) (criticizing this approach); see also Coyle, Canons of Construction, supra note 153, at 642–47 (discussing the distinction between internal law and whole law).
  168. . Siegelman, 221 F.2d at 194.
  169. . Restatement (Second) of Conflict of Laws § 187(3) (Am. Law Inst. 1971); id. at cmt. h.
  170. . Id. § 187, cmt. h.
  171. . First Wis. Nat’l Bank v. Nicolaou, 270 N.W.2d 582, 585 (Wis. Ct. App. 1978) (“The proper construction of the choice of law provision . . . is that it makes Wisconsin internal law, i.e., the law which would govern purely domestic Wisconsin cases, applicable to any act of enforcement of the contract wherever it may occur.”).
  172. . Carlos v. Philips Bus. Sys., Inc., 556 F. Supp. 769, 774 (E.D.N.Y. 1983).
  173. . McLaughlin v. Reynolds, 886 F. Supp. 902, 904 (D. Me. 1995) (“without regard to its conflict of laws rules”).
  174. . Packquisition Corp. v. Packard Press New Eng., No. 91-5966, 1992 U.S. Dist. LEXIS 17460, at *7 (E.D. Pa. Nov 1, 1992).
  175. . First Nat’l Bank of Mount Dora v. Shawmut Bank of Bos., 389 N.E.2d 1002 (Mass. 1979); see also Wiesenberger Servs., Inc. v. Response Analysis Corp., 365 F. Supp. 258, 258 (S.D.N.Y. 1973) (“This agreement shall be governed by, and construed in accordance with the laws of the State of New York, other than conflicts of law rules.”).
  176. . This upward trend appears to have continued in the first two decades of the twentieth century. A review of 351 choice-of-law clauses written into bond indentures filed with the Securities Exchange Commission in 2016 found that 55 percent contained language excluding conflict-of-laws rules. Coyle, U.S. Bond Indentures, supra note 162. A separate review of 159 international supply agreements filed with the SEC between 2011 and 2015 found that approximately 78 percent contained language excluding conflict-of-laws rules. Coyle & Weidemaier, supra note 162.
  177. . 600 Grant St. Assocs. Ltd. v. Leon-Dielmann Inv. P’ship, 681 F. Supp. 1062, 1064 (S.D.N.Y. 1988).
  178. . Brinderson-Newberg Joint Venture v. Pac. Erectors, Inc., 690 F. Supp. 891, 892 (C.D. Cal. 1988) (“Except as provided in the General Contract as incorporated herein, this subcontract shall be interpreted and governed substantively and procedurally, including periods for limitations of actions, by the law of the State of California.”).
  179. . Clauses specifically referencing statutes of limitations are not included on the chart because there were only two such clauses in the data set.
  180. . It is not altogether clear that drafters are fully cognizant of the significance of the word “enforced” when they write it into their clauses, which in turn makes it hard to assess the significance of clauses that use this term. This fact notwithstanding, a number of courts have held that the use of the word “enforced” in a choice-of-law clause connotes an intent to select the procedural law of the chosen jurisdiction. See Coyle, Canons of Construction, supra note 153, at 655 n.113 (collecting cases).
  181. . Stark v. Nw. Nat’l. Life Ins., 167 F. 191, 192 (D. Minn. 1909) (contract stating that “the laws of the state of Minnesota . . . constitute and form a part of each . . . policy”).
  182. . See Cotter v. Lyft, Inc., 60 F. Supp. 3d 1059, 1066 (N.D. Cal. 2014); Steven W. Feldman, Statutes and Rules of Law as Implied Contract Terms: The Divergent Approaches and a Proposed Solution, 19 U. Pa. J. Bus. L. 809, 852 (2017).
  183. . See Alaska Packers Ass’n v. Indus. Accident Comm’n, 1 Cal. 2d 250 (1934) (“[T]he parties . . . hereby agree, as an incident of this contract of employment, to accept and be bound by the provisions of the said Workmen’s Compensation Act of Alaska for any and all injuries arising out of and in the course of their employment, and further agree to accept as their exclusive remedy for any and all industrial injuries, the provisions of the said Workmen’s Compensation Act of Alaska.”).
  184. . Only a smattering of clauses expressly referenced a federal treaty. Vision Graphics, Inc. v. E.I. Du Pont de Nemours & Co., 41 F. Supp. 2d 93, 97 (D. Mass. 1999) (“THIS AGREEMENT WILL BE GOVERNED BY AND CONSTRUED IN ACCORDANCE WITH THE LAWS OF THE STATE OF DELAWARE without regard to its choice of law provisions (and the United Nations Convention on the International Sale of Goods shall specifically not apply).”).
  185. . U.S. Fid. & Guar. Co. v. Nw. Eng’g Co., 112 So. 580, 581 (Miss. 1927) (involving a contract stipulating that the rights of the parties were to be “governed by the laws of Wisconsin existing at the time of making the contract”).
  186. . Stark, 167 F. at 192 (involving a contract stipulating that it was to be governed by “the laws of the state of Minnesota as they now exist or as may hereafter be amended”).
  187. . See Vitek Danilowicz, ‘Floating’ Choice-of-Law Clauses and Their Enforceability, 20 Int’l Law. 1005 (1986).
  188. . See English Co. v. Nw. Envirocon, 663 N.E.2d 448, 451 (Ill. App. Ct. 1996) (“The Agreement shall be tried or heard in and governed by and construed in accordance with the governing laws of the home state of the initial defendant to any action at law brought by or on behalf of the other party, including countersuit by the initial defendant.”).
  189. . Dean Witter Reynolds, Inc. v. Espada, 959 F. Supp. 73, 75 (D.P.R. 1997) (“The law of the State of New York will apply in all respects, including but not limited to determining of applicable statutes of limitation and available remedies.”).
  190. . Cf. Espresso Disposition Corp. 1 v. Santana Sales & Mktg. Grp., Inc., 105 So. 3d 592, 594 (Fla. Dist. Ct. App. 2013).
  191. . See Michael I. Meyerson, The Efficient Consumer Form Contract: Law and Economics Meets the Real World, 24 Ga. L. Rev. 583, 605 (1990) (“Intuitively, any profit-maximizing business would prefer to shift a risk to the other party if it could do so at no additional cost. Because consumers lack the knowledge to evaluate the cost of the risk, a rational seller will draft contract terms that shift risks to the consumer.”).
  192. . See Robert A. Hillman & Jeffrey J. Rachlinski, Standard-Form Contracting in the Electronic Age, 77 N.Y.U. L. Rev. 429, 452 (2002) (“This narrow cognitive focus that people bring to complex decisions creates a temptation for businesses to offer enticing prices and terms concerning the negotiable portions of the form and to make up for any concessions by drafting one-sided boilerplate terms. Consumers will focus their cognitive skills on the ‘important’ terms, such as price, but ignore the hidden costs buried in the boilerplate.”).
  193. . Russell Korobkin, Bounded Rationality, Standard Form Contracts, and Unconscionability, 70 U. Chi. L. Rev. 1203, 1244 (2003) (“[A]lthough the market should be expected to provide efficient salient contract terms to the advantage of buyers as a class and sellers as a class, no such assumption about non-salient terms is defensible.”).
  194. . In one sense, it may be rational for some lawyers to invest relatively little time in updating and revising their choice-of-law clauses. Clients are generally uninterested in paying their attorneys to research and revise contract boilerplate, and attorneys (rationally) tend to devote their time to substantive contract issues that are of greater concern to their clients.
  195. . Even if these companies were concerned about whether these clauses would be enforced, there was no penalty for including them in the agreement. At worst, the judge would ignore the clause and apply the law required under the forum’s conflict rules.
  196. . Path dependence likely goes a long way towards explaining the stickiness of contract language. In the absence of any pressing reason to update a contract, the path of least resistance is to leave it be. This is particularly true when the provision at issue is of marginal interest to nonlawyers and is typically found at the very end of the contract along with other “miscellaneous” provisions.
  197. . See Wendy Netter Epstein, Public-Private Contracting and the Reciprocity Norm, 64 Am. U. L. Rev. 1, 35 (2014) (observing that “the positive reciprocity norm—meaning that people reward kind actions—has been shown to often constrain actors’ behavior, resulting in deviations from what the rational actor model would predict”).
  198. . Stephen Choi, Mitu Gulati & Eric A. Posner, The Dynamics of Contract Evolution, 88 N.Y.U. L. Rev. 1, 9–10 (2013) [hereinafter Choi et al., Dynamics]; see also Stephen J. Choi, Mitu Gulati & Eric Posner, The Evolution of Contractual Terms in Sovereign Bonds, 4 J. of Legal Analysis 131, 152 (2012) (discussing the evolution of other clauses in sovereign debt agreements).
  199. . Choi et al., Dynamics, supra note 197, at 27 (“[S]hifts in boilerplate contract terms do not occur without some initial shock.”).
  200. . Florencia Marotta-Wurgler & Robert Taylor, Set in Stone? Change and Innovation in Consumer Standard-Form Contracts, 88 N.Y.U. L. Rev. 240, 247–48 (2013) (describing shocks as “changes in legal interpretations of terms, or technological advances”).
  201. . Choi et al, Dynamics, supra note 197, at 37 (“[T]op market participants switch from being defenders of the status quo to promoters of their own individual visions of the anticipated new standard.”).
  202. . See Stephen J. Choi & Mitu Gulati, Contract as Statute, 104 Mich. L. Rev. 1129, 1133, 1157 (2006) (“[W]ith boilerplate clauses, dispersed market participants may lack the ability to coordinate, at least initially, to clarify the language in subsequently adopted terms.”).
  203. . Mitu Gulati & Robert E. Scott, The Three and a Half Minute Transaction: Boilerplate and the Limits of Contract Design 60 fig.2 (2013).
  204. . Id. at 60 figs. 2 & 3.
  205. . Christopher Drahozal & Peter Rutledge, “Sticky” Arbitration Clauses? The Use of Arbitration Clauses After Concepcion and Amex, 67 Vand. L. Rev. 955, 982–83 (2014) (invoking shock model to explain changes to arbitration clauses).
  206. . Matthew Jennejohn, The Architecture of Contract Innovation, 59 B.C. L. Rev. 71, 89 n.83 (2018).
  207. . David Hoffman, Whither Bespoke Procedure?, 2014 U. Ill. L. Rev. 389, 394 (2014).
  208. . With respect to the pace of change, the shock model presupposes that the process of contractual change begins with something akin to a lightning bolt. In the choice-of-law clause context, however, a better metaphor is that of a pebble perched on the top of the mountain. If the pebble falls just the right way, it will dislodge additional pebbles, which will in turn dislodge still more pebbles, until the avalanche begins in earnest.
  209. . This list of search terms included (1) “choice of law provision,” (2) “choice of law clause,” (3) “in accordance with the laws of the state” w/30 contract, (4) “construed in accordance with the laws of the state,” (5) “governed by” w/20 “according to the,” (6) “construed” w/5 “according to the laws,” (7) “governed by and construed,” (8) “the laws of the state of” w/20 “provision,” (9) “the laws of the state of” w/20 “interpreted,” (10) “the law of the state” w/20 “interpreted,” (11) “construed” w/5 “interpreted,” (12) “construed” w/5 “governed by,” (13) “interpreted” w/5 “governed by,” (14) “construed solely,” (15) “determined in accordance with the laws” w/30 contract, (16) “determined in accordance with the laws” w/30 “agreement,” (17) “in accordance with the laws of” w/30 “agreement,” (18) “governed” w/5 “interpreted,” (19) “governed” w/5 “construed,” (20) “construed” w/5 “interpreted,” (21) “shall be deemed to have been made,” (22) validity w/10 “shall be determined,” (23) “shall be governed as to validity,” (24) “the place of the contract,” (25) agreement w/5 “subject to,” (26) “Without regard to” w/5 conflict!, (27) construction w/10 interpretation w/10 agreement, (28) construction w/10 validity w/10 agreement, (29) “this agreement and its enforcement,” (30) “agreement” w/8 “deemed” w/8 “made,” and (31) “agreement” w/8 “deemed” w/8 “executed.”
  210. . See, e.g., David A. Hoffman, Alan J. Izenman & Jeffrey R. Lidicker, Docketology, District Courts, and Doctrine, 85 Wash. U. L. Rev. 681 (2007); Ahmed E. Taha, Data and Selection Bias: A Case Study, 75 UMKC L. Rev. 171 (2006).