Article Thirty-nine of the Magna Carta (1215) proclaimed that “no free man shall be taken or imprisoned or disseised or outlawed or exiled or in any way ruined, nor will we go or send against him, except by the lawful judgment of his peers or by the law of the land.” This “law of the land” requirement, which is often called the principle of legality, prohibited unilateral, arbitrary action by the king against certain protected private interests. Executive and judicial deprivations of such interests could take place only pursuant to valid legal authority.

The phrase “due process of law” made its first appearance in a statute of 1354 concerning court procedures. “Due process of law” meant that judgments could issue only when the defendant was personally given the opportunity to appear in court pursuant to an appropriate writ (i.e., was served process). The phrase retained this technical meaning in English law into the eighteenth century.

At the time of the drafting of the Bill of Rights, at least eight state constitutions contained clauses restraining government from depriving persons of life, liberty, or property except pursuant to the law of the land. The Fifth Amendment, which otherwise tracked the form of these state provisions, used the phrase “due process of law” instead of “law of the land.” The reasons for this change in terminology are uncertain, but it is likely that the founding generation was misled by some seventeenth-century statements of Sir Edward Coke (familiar to virtually all the Founders), who in 1642 had declared—wrongly, in the judgment of modern historians—that the phrases “law of the land” and “due process of law” were essentially equivalent. Accordingly, the constitutional meaning of “due process of law,” as understood in America in the late eighteenth century, almost certainly refers to the principle of legality rather than to pleading technicalities.

Until very close to the time of the framing, the judicial power was generally viewed as an aspect of executive power. In eighteenth-century America, therefore, the phrase “without due process of law” meant something like “executive or judicial action taken without lawful authorization or not in accordance with traditional procedural forms of justice.” The Supreme Court modestly extended the principle to Congress in Murray’s Lessee v. Hoboken Land & Improvement Co. (1856)—the first Supreme Court case that turned on the meaning of the Due Process Clause. There the Court determined that the Due Process Clause limited the power of Congress to authorize novel forms of adjudication. The case involved a constitutional challenge to a statutory procedure in which the government collected deficiencies from tax collectors without first having a court determine whether the tax collector really owed the money to the government. The Court found that the clause “is a restraint on the legislative as well as on the executive and judicial powers of the government, and cannot be so construed as to leave congress free to make any process ‘due process of law,’ by its mere will.” In order to determine whether legislatively prescribed forms of adjudication violated due process of law, the Court looked to “those settled usages and modes of proceeding existing in the common and statute law of England.” The Court found a long tradition in English and American law of auditing tax collectors without prior hearings and accordingly upheld the practice.

The Due Process Clause requires that deprivations of “life, liberty, or property” be accompanied by due process of law. The deprivation of other interests not enumerated here need not be accompanied by due process of law. When the Due Process Clause was ratified in 1791, the meaning of “liberty” as a personal right was clear. Sir William Blackstone, whose influence on the founding generation was enormous, wrote in his Commentaries on the Laws of England (1765–1769) that the right to liberty meant “the power of locomotion, of changing situation, or removing one’s person to whatsoever place one’s own inclination may direct; without imprisonment or restraint, unless by due course of law.” That definition is quite narrow and focused, and it excludes such matters as bodily integrity and reputation. Those interests, however, were encompassed by Blackstone’s (expansive to modern ears) definition of “life,” which referred to an array of rights lumped together under the general heading of personal security: “a person’s legal and uninterrupted enjoyment of his life, his limbs, his body, his health, and his reputation.” The term “property” in 1791 was more ambiguous. It could have referred to land, to land plus chattels, to anything of exchangeable value, or (what seems most likely) to whatever interests common law courts would have recognized as property entitled to legal protection. None of these understandings would include as property future enjoyment of government benefits, such as jobs or licenses. The law sharply distinguished between property rights and mere privileges that the government could continue or terminate at its pleasure.

The eighteenth-century lawyer trying to define the phrase “life, liberty, or property” would have faced a sticky problem that could not be answered by reference to Blackstone, tradition, or any other authority: Do these terms draw their meaning from federal law, state law, or both? Does the Constitution contain its own internal definitions of those terms, so that the Constitution itself determines whether a particular interest is property? Are they defined by reference to the laws of the states, so that a particular interest might be constitutional “property” in Pennsylvania but not in New York? Or must the universe of constitutionally protected interests be determined by some combination of federal and state law?

Perhaps the best answer, though it is impossible to say decisively, is that federal law sets the outer boundaries of “life, liberty, and property” and state law fixes the details. Surely the Constitution does not itself determine whether a particular estate in land, such as a surface estate on mining land, is or is not “property” for purposes of the Due Process Clause or whether water rights must be appropriative or riparian; either answer is permissible and, accordingly, can vary from one jurisdiction to another. But if a state decided that land itself was henceforth no longer to be considered “property,” that would pass the boundaries of acceptability. Federal law thus establishes for each term a “core” of meaning that no jurisdiction can alter, but beyond that core, governments are free to expand or contract the range of constitutionally protected interests.

Accordingly, the original meaning of the Due Process Clause was essentially that the federal government could not take people’s lives, health, reputation, freedom of movement, or common law property without prior legal authorization and without following traditional judicial procedures. Describing modern doctrine is considerably more difficult.

Modern doctrine has significantly modified the original understanding of how one determines compliance with “due process of law.” Instead of reference to traditionally accepted procedural forms, contemporary law dating back more than a century has judged the adequacy of procedures by a mélange of practical factors that resist easy reduction. Late nineteenth-century and early twentieth-century cases described due process as procedures “appropriate to the case, and just to the parties to be affected,” Hagar v. Reclamation Dist. No. 108 (1884), and the absence of due process as procedures that are “inadequate or manifestly unfair.” ICC v. Louisville & Nashville R.R. Co. (1913). As Justice Felix Frankfurter summarized matters in a famous concurring opinion in Joint Anti-Fascist Refugee Committee v. McGrath (1951), due process “is compounded of history, reason, the past course of decisions, and stout confidence in the strength of the democratic faith which we profess. . . . It is a delicate process of judgment by those whom the Constitution entrusted with the unfolding of the process.”

In Mathews v. Eldridge (1976), the Court clarified that this “delicate process of judgment” generally requires consideration of

first, the [significance of the] private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the [weight of the] Government’s interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail.

While there is no indication that the Court in Mathews thought that it was changing or limiting the previous century’s focus on fundamental fairness as determined by all-things-considered reasoning, within a few years this formulation from Mathews took on a life of its own, and today it serves as the near-exclusive frame of reference for analyzing procedural adequacy under the Due Process Clause. This modern balancing-of-factors approach is universally criticized as unpredictable. Many observers also object to the Court’s optimistic goal of accurate decision-making, and there is considerable disagreement about which other possible goals of procedure should be factored into the mix. For example, if one limits the due process analysis to the concerns identified in Mathews, then the interests of children and custodial parents would not seem to be relevant to determining the procedures for enforcement of child support, which struck the Court’s originalist-leaning justices in 2011 as implausible. See Turner v. Rogers (2011). Nonetheless, a relatively rote application of the Mathews factors is the standard modern approach to establishing the constitutional adequacy of procedures.

One of the trickiest questions in modern law concerns the timing of procedures: which procedures (if any) must come before the government deprives people of protected interests? The law in this area remains unsettled in many important respects and defies simple description. Perhaps the most that can be said—and can be said only as a rough generalization—is that there is a presumption in favor of pre-deprivation procedures that can be overcome when the value of such procedures would be low, the potential risks to the government or public from delaying the deprivation would be high, or the person deprived would have an adequate remedy after the deprivation.

The most dramatic transformations in modern due process have concerned the range of interests encompassed by the phrase “life, liberty, or property.” As late as 1950, the original meaning still largely held sway, though Blackstone’s broad understanding of “life” as including bodily integrity and reputation mysteriously vanished in favor of a much narrower meaning. No doubt this development, which was never expressly acknowledged, put pressure on the other terms in the enumeration (“liberty” and “property”) to include such worthy interests as physical integrity and reputation. More importantly, the rise of the post–New Deal administrative state vastly expanded the range of circumstances under which official action could affect people’s lives; and the concomitant expansion of government benefits, jobs, and licenses raised the stakes of excluding such interests from procedural protection. By the early 1960s, a majority of the Court was prepared to treat the phrase “life, liberty, or property” as a convenient shorthand for any interest whose loss would be grievous rather than as a list of three distinct terms with distinct, ascertainable meanings—a development that some commentators half-jokingly described as the emergence of “lifelibertyproperty.” On this new understanding, government benefits could easily constitute an interest whose loss would be grievous and whose deprivation therefore required some measure of procedure.

In 1970, the Court formalized this understanding in Goldberg v. Kelly, where the State of New York, in its argument to the Supreme Court concerning hearings prior to termination of benefits under the Aid to Families with Dependent Children Act, did not even argue that expected future receipt of AFDC benefits was not a constitutionally protected interest. Subsequent cases quickly extended constitutional protection to such interests as government licenses and reputation.

In 1972, the Supreme Court established the framework of modern law in Board of Regents of State Colleges v. Roth. That case reestablished some differentiations among the three enumerated categories of protected interests. The Court held that “liberty” and “property” were distinct terms with ascertainable meanings, though “life” continues to be conspicuously absent from modern recitations of the range of protected interests in anything other than its narrowest and most literal application. The Court explicitly stated, however, that these terms would not be construed in accordance with their original meaning, but would have to be construed to include the extended range of interests recognized in prior case law, including government benefits.

Accordingly, the Court has expanded the definition of the term “liberty,” drawing upon Meyer v. Nebraska (1923), in which it declared that “liberty” includes “not merely freedom from bodily restraint but also the right of the individual to contract, to engage in any of the common occupations of life, to acquire useful knowledge, to marry, establish a home and bring up children, to worship God according to the dictates of his own conscience, and generally to enjoy those privileges long recognized . . . as essential to the orderly pursuit of happiness by free men.” The Court’s extraordinary expansion of the concept of “liberty” reached its apogee in the famous (some critics say infamous) declaration by Justice Anthony Kennedy: “At the heart of liberty is the right to define one’s own concept of existence, of meaning, of the universe, and of the mystery of human life.” Planned Parenthood of Southeastern Pennsylvania v. Casey (1992). He quoted this passage a decade later, holding in Lawrence v. Texas (2003) that the state could not prohibit homosexual sodomy, and in United States v. Windsor (2013), that the federal government could not deny federal benefits to legally married same-sex couples.

Although one may have the right to define reality in any way that one may please, acting upon those beliefs is another story. So despite Justice Kennedy’s sweeping language, the constitutional meaning of liberty has been held not to include, at least under some circumstances, a right to government employment, an interest in reputation, or many interests claimed by prisoners. The government is free to construct these excluded interests to be constitutionally protected through statutes and regulations by specifying a clear causal connection between satisfaction of criteria of eligibility and receipt of a benefit, but they are not automatically protected as a matter of constitutional command.

Today, it is state law that primarily defines the term “property.” Interests within the traditional common-law understanding of property are generally still considered to be property. Different states can, for example, prescribe different kinds of estates in land as “property,” but a state could not by statutory or judicial fiat declare land per se (or money in a bank account) not to be private property for purposes of the Constitution. Interests beyond the traditional common law understanding, such as government benefits and licenses, are constitutionally protected if statutory or regulatory provisions draw a clear causal line from the satisfaction of eligibility criteria to the receipt of benefits. The case law distinguishes the substance of the created interest from the procedures for its termination. The latter is what the Due Process Clause protects. Within the zone beyond the constitutional core of “liberty” and “property,” government can determine which substantive interests shall receive due process protection, but once that substantive decision is made, the constitutional law of due process assesses the adequacy of the procedures. In other words, the government may not make acceptance of “unconstitutional” termination procedures a condition of receiving government benefits.

If an interest does not fall within the meaning of the phrase “life, liberty, or property,” the Due Process Clause does not mandate any particular procedures for its deprivation. Other sources of law, whether constitutional or statutory, may well do so, but the Due Process Clause is, so to speak, “turned off.” There are several other “on-off switches” that also determine the applicability of the Due Process Clause.

First, the clause applies only to government action; private entities are not bound by the Fifth Amendment or, indeed, by anything in the Constitution except the Thirteenth Amendment. This can pose difficult questions when the acting entity is nominally private but is involved in some fashion with the government. See State Action Clause (Amendment XIV, Section 1). Second, modern law holds that the word “deprived” in the Due Process Clause means an intentional (or, at a minimum, a reckless) taking of a protected interest. Losses inflicted by government negligence do not implicate the Due Process Clause.

Third, and most importantly, administrative agencies are responsible for the vast bulk of governmental actions that work deprivations of interests within the compass of the Due Process Clause, but large classes of agency action have been held to fall outside the clause’s protection. Agencies engage in two forms of official action: rule-making, which strongly resembles in form and function the promulgation of a statute by the legislature, and adjudication, which strongly resembles in form and function the decision of a case by a court. The Due Process Clause has never been understood to impose procedural requirements on legislatures (though it does, under modern understandings, regulate the content of legislation that authorizes executive or judicial procedures). For almost a century, courts have held that agency rule-making shares in this legislative immunity from due-process analysis; agency rule-making is subject to no constitutional procedural requirements. Bi-Metallic Investment Co. v. State Board of Equalization of Colorado (1915). Agency adjudication, however, is subject to due process analysis, but agencies do not stand in the same shoes as courts. Procedures that would be obviously inadequate in judicial proceedings are considered constitutionally adequate for agency adjudication. The size of the gap is uncertain, which typifies the complexity of the modern law of procedural due process.