THE CASE AGAINST CIVIL EX POST FACTO LAWS

Vol. 15 No. 2-3

Steve Selinger

According to Article I, Section 10 of the Constitution, "no State shall pass any ex post facto Law.'' A similar provision that applies to Congress is found in Section 9 of the same article. At first glance these constitutional prohibitions seem simple enough--retroactive laws violate the Constitution. Unfortunately, the issue is not so simple. With one ruling in 1798, the Supreme Court succeeded in muddling the issue of ex post facto laws by holding that the prohibition of retroactive laws applies only to criminal, not civil, laws.

In The Constitution of Liberty, F. A. Hayek (1960: 205-20) notes that some coercion, while unavoidable in a civil society, can be minimized by requiring that coercive actions comply with general rules that are known in advance by individuals. If individuals know the law, they can base their actions upon established rules and minimize the ill effects of coercion. Hayek states that not all legislative enactments will satisfy the three criteria of what he calls "true law''--generality, certainty, and equality. He argues that true law provides the general rules which minimize coercion and that legislative enactments which do not satisfy these criteria are objectionable. He writes that the law must be general, that it must be known and certain, and that it should apply equally to all. A necessary condition for the law to be known and certain is a prohibition on ex post facto laws. After all, the law can hardly be known and certain if new laws can be made to apply retroactively to actions already performed.

From a policy standpoint, as Hayek's analysis indicates, ex post facto laws are riddled with problems. Unfair and unpredictable, ex post facto or retroactive laws mar the American legal system and create an abundance of problems.[1]

This article examines the origins of the constitutional interpretation of ex post facto laws, reviews sources favoring the prohibition of ex post facto civil laws, considers the damaging impact of retroactive laws on property rights, and proposes a solution to the debate over retroactive laws in which ex post facto civil as well as criminal laws would be constitutionally prohibited unless just compensation is provided for unfair retroactivity. Given that the clear constitutional ban on ex post facto laws does not distinguish between criminal and civil laws and given our own intuitions about fairness in the legal system, the prohibition on ex post facto laws should be extended to civil laws in order to prevent unfair and capricious changes in the law.

Calder v. Bull: Origins of the Distinction between
Civil and Criminal Ex Post Facto Laws

The Supreme Court first held that the constitutional prohibition against ex post facto laws applied only to criminal laws in the landmark opinion of Calder v. Bull (1798). The issue in the case, which arose from the Supreme Court of Connecticut, was whether the act of the Connecticut legislature to set aside a decree of a probate court (which had the effect of divesting the appellants of certain property) was an ex post facto law.

The matter involved rights obtained under a will. Calder had obtained the right to possession of certain property via a probate court ruling, from which there was no right of appeal. The Connecticut legislature then passed a law setting aside the decree of the probate court and allowing a new hearing and a right of appeal. On rehearing, the probate court found in favor of Bull. On Calder's appeal to the Supreme Court, he argued that the law passed by the Connecticut legislature was an ex post facto law and therefore prohibited by Article 1, Section 10 of the Constitution, which states in part that no state shall pass any "ex post facto law.''

Surprisingly, the Supreme Court rejected Calder's approach, holding that the constitutional prohibition of ex post facto laws applied only to criminal laws. In an opinion written seriatim by Justices Samuel Chase, William Paterson, and James Iredell, Justice Chase acknowledged that the right to certain property had been awarded to Calder as a result of the probate court decision. However, due to a subsequent law, the new hearing of the matter in the probate court, and the resulting decision, Calder's right to recover that property was negated and the property rights were awarded to Bull. According to Justice Chase, the law that took away Calder's right to that property was not an ex post facto law. Ironically, Justice Chase also acknowledged that there are certain laws that a legislature may not constitutionally enact, among them being "[a] law that takes property from A and gives it to B'' (Calder: 388). Although it seems obvious that this language applies to the facts of Calder, the Court did not view it that way.

In his reasoning, Justice Chase cited Article I, Section 10 of the Constitution, which contains the ex post facto prohibition, and pointed out that it also prohibits: (1) the making of anything but gold and silver coin a tender in the payment of debts, and (2) the passage of any law impairing the obligations of contracts. According to Justice Chase, those two provisions were inserted to secure private rights. Chase reasoned that if the prohibition against making ex post facto laws had been intended to secure personal rights from being affected or injured by such laws, and the prohibition against ex post facto laws was sufficiently extensive for that object, then the listing of the coin and contract restraints was redundant.

Justices Paterson and Iredell concurred with Justice Chase. Justice Paterson argued that the bar on the impairment of contracts suggested that the framers of the Constitution intended the prohibition on ex post facto laws to include "crimes, pains, and penalties, and no further.'' According to Justice Iredell, the purpose of the ex post facto clause did not extend to "civil cases, to cases that merely affect the private property of citizens'' (Calder: 400).

Strangely enough, while Calder has been discussed at length by legal historians, the essential logic of the Court's decision appears to have been ignored. Most discussions of the case have been historical inquiries concerning how "ex post facto'' was used when the Constitution was first written. Such inquiries have shown that the prohibition on ex post facto laws was probably not intended to be restricted to criminal laws (e.g., Field 1920, Smead 1936). Perhaps the most thorough treatment of ex post facto laws can be found in William Crosskey's Politics and the Constitution (1953). Although Crosskey criticizes the Calder decision, he does not appear to address directly the Court's chief argument.

According to Crosskey (1953: 351), following the Calder decision, Supreme Court Justice William Johnson sought to persuade a later Court that Calder had been wrongly decided. In dissecting and refuting the reasons for the Calder decision, Johnson wrote in Satterlee v. Matthewson (1829):

It is obvious, in the case of Calder v. Bull, that the great reason which influenced the opinion of the three judges who gave an exposition of the phrase "ex post facto'' was that they considered its application to civil cases as unnecessary, and fully supplied by the prohibition to pass laws impairing the obligation of contracts.

Justice Johnson went on to assert that the class of civil ex post facto laws was broader than the simple class of laws impairing the obligation of contracts. As such, he answered that it would not suffice to equate the prohibition against civil ex post facto laws with the prohibition against laws impairing contracts. According to Justice Johnson,

The learned judges could not then have foreseen the great variety of forms in which the violation of private right have since been presented to this Court. The case of a legislature declaring a void deed to be a valid deed, is a striking one to show, both that the prohibition to pass laws violating the obligation of contracts is not a sufficient protection to private rights; and that the policy and reason of the prohibition to pass ex post facto laws does indeed extend to civil as well as criminal cases [Satterlee: 685].

Unfortunately, Justice Johnson's reply to the Calder opinion misinterprets the Court's argument. The Court in Calder did not argue that civil ex post facto laws need not be prohibited because of the existence of the Contracts Clause, as Johnson interprets them as saying. The majority did not say or assume that the prohibition against civil ex post facto laws was unnecessary because the prohibition against laws impairing contracts covered all possible civil ex post facto laws.

Rather, the Calder majority argued that the prohibition on ex post facto laws could not have been meant to apply to civil laws because the addition of the prohibition on laws impairing contracts would have been superfluous. Since laws impairing contracts are a subclass of civil ex post facto laws, if all civil ex post facto laws are outlawed, it is unnecessary to further outlaw this subclass of ex post facto laws. Hence, the Calder Court reasoned, the prohibition cannot have been meant to apply to all civil ex post facto laws. It is this argument, and not the one that Johnson misinterprets the Court as making, that must be addressed.

The Calder majority's argument is easily refuted with the reasoning that there is nothing inconsistent with interpreting the ex post facto prohibition as applying to all laws, including civil laws. For example, Article I, Section 10 of the Constitution prohibits states from making legal tender out of anything but gold and silver coins and there is a constitutional prohibition against laws impairing contracts. These examples are illustrative of the prohibition against ex post facto laws, but do not encompass all possible ex post facto laws. If two individuals have a binding contract, and a legislature then passes a law abrogating the contract, an ex post facto law has been applied to the contract. Similarly, if two individuals make an agreement that a debt shall be paid in gold or silver, and a legislature passes a law saying that newly created paper money shall suffice for the payment of the debt, an ex post facto law has been applied to them.

An action can be prohibited by both a general and a more specific clause of the Constitution. For instance, a law substituting paper money for gold and silver tender would be prohibited by both the contracts clause and the gold and silver clause. Strictly speaking, the clause in the Constitution forbidding laws that impair contracts by substituting other legal tender in place of gold and silver is unnecessary as it is covered by the more general contracts clause. Yet no one would argue that the contracts clause should not apply to other laws that seek to change standards for legal tender and impair contracts. For instance, a law impairing contracts to pay in platinum would still be unconstitutional even though only gold and silver are mentioned in the legal tender clause; the inconsistency with the more general contracts clause would still be sufficient to render the law unconstitutional.

Similarly, while a law impairing contracts is one specific type of ex post facto law, as the Court correctly pointed out, it is not the only type of ex post facto law. Interpreted in this way, there is nothing inconsistent in applying the prohibition against ex post facto laws to other civil laws. The other two clauses mentioned in the Constitution are merely examples of more specific types of ex post facto laws that are prohibited.

The Court seems to have been guided in its logic by the maxim that the "specific controls the general,'' a maxim that is enshrined in California Civil Code st3354, which states, "[p]articular expressions qualify those which are general.'' It is true that if there is a conflict between a more specific statute and a general one, the specific one controls. Where there is no conflict, however, a general provision is still applicable (as the examples of the contracts clause and legal tender clause show).

This position is buttressed by the fact that the Constitution does not explicitly prohibit Congress from passing laws impairing the obligation of contracts. Regarding Congress, the Constitution merely states that "Congress shall pass no bills of attainder or ex post facto laws.'' It is in the subsequent section, regarding prohibitions on laws the states may pass, that the more specific clauses referencing the prohibition on laws impairing the obligation of contracts and legal tender are found.

It is thus problematic to base an argument against the application of the ex post facto clause to civil law on the existence of the contract and legal tender clauses (as the Calder Court does), because those clauses do not even appear in the section of the Constitution restricting laws that Congress can make. By the logic in Calder it would follow that Congress is not free to pass civil ex post facto laws. Yet this logic gives us the strange result that states can pass civil ex post facto laws (because of the existence of the contract and legal tender clauses) while Congress cannot pass civil ex post facto laws (because the prohibition on ex post facto laws with respect to Congress is not accompanied by the contract and legal tender clauses).

The second argument that Justice Chase makes in Calder is that state constitutions constructed before the U.S. Constitution explicitly prohibited criminal ex post facto laws (Calder: 391). The Massachusetts Constitution, for instance, refers to "laws made to punish actions done before the existence of such laws,'' while the Maryland Constitution refers to "laws punishing facts committed before the existence of such laws.''

William Crosskey (1953: 342-51) convincingly shows that Justice Chase's argument in Calder about state constitutions is flawed. Justice Chase quotes from some of the referenced state constitutions, misquotes others, and ignores the New Hampshire Constitution, which expressly applied a prohibition against ex post facto laws to both civil and criminal laws. Crosskey notes the various state and federal circuit-court cases that, until Calder, had interpreted the prohibition as applying to civil as well as criminal laws.

Although Crosskey's historical arguments are persuasive, it is also worth noting that the Court's arguments in Calder are logically problematic. Even if we grant the Court's position that the majority of state constitutions at the time did reference ex post facto laws in an explicitly criminal context, that fact fails to prove the Calder majority's contention that the prohibition of ex post facto laws in the U.S. Constitution was intended to apply only to criminal laws. Unlike ex post facto clauses in some state constitutions, the one in the U.S. Constitution does not contain an explicit reference to criminal laws.

In that case, the legislatures of the time chose not to limit the application of the ex post facto clause to criminal acts through explicit language, although they clearly knew how to create an express limitation. The Court should not presume to insert such language when a legislature (or constitutional convention) has chosen to withhold the language. It thus seems that the Court's own remarks about the state constitutions of the time (even if such constitutions had been correctly and fully quoted) do not justify its conclusion that ex post facto laws apply only in a criminal context. If anything, they serve to prove just the opposite point.

Sources Favoring the Prohibition of Civil Ex Post Facto Laws

The Federalist Papers: Favoring Regular Course of Business

Beyond the flawed logic of Calder, there are many sources of support for the prohibition of civil ex post facto laws. The Federalist Papers, written in support of the Constitution, may provide guidance as to the intended meaning of the term "ex post facto'' in the Constitution. The Federalist Papers refer to ex post facto laws in both the civil and criminal context. The authors intended for the prohibition of ex post facto laws to extend to civil laws as well as criminal laws, as James Madison noted in Federalist No. 44:

Bills of attainder, ex-post facto laws, and laws impairing the obligation of contracts are contrary to the first principles of the social compact, and to every principle of sound legislation.... [T]he sober people of America are weary of the fluctuating policy which has directed the public councils. They have seen with regret and indignation that sudden changes and legislative interferences, in cases affecting personal rights, become jobs in the hands of enterprising and influential speculators, and snares to the more industrious and less informed parts of the community. They have seen, too, that one legislative interference is but the first link of a long chain of repetitions, every subsequent interference being naturally produced by the effects of the preceding. They very rightly infer, therefore, that some thorough reform is wanting, which will banish speculations on public measure, inspire a general prudence and industry, and give a regular course to the business of society.

The references to "prudence and industry'' and giving a "regular course to the business of society'' suggest that the constitutional prohibition against ex post facto laws should be seen as applying to both civil and criminal laws because of public policy reasons. In Federalist No. 84, Madison also refers to criminal laws:

The establishment of the writ of habeas corpus, the prohibition of ex-post-facto laws, and of titles of nobility, to which we have no corresponding provision in our Constitution, are perhaps greater security to liberty and republicanism than any it contains. The creation of crimes after the commission of the fact, or ... the subjecting of men to punishment for things which, when they were done, were breaches of no law ... have been, in all ages, the favorite and most formidable instruments of tyranny.

Taken together, these passages strongly suggest that the authors of the Federalist Papers intended the prohibition against ex post facto laws to apply to both civil and criminal laws.

The Obscure Line Between Criminal and Civil Laws

An additional concern regarding the retroactive application of civil laws is that the distinction between civil and criminal laws is frequently discretionary. For example, both the Internal Revenue Service and the Securities and Exchange Commission can often choose to file suit on either a civil or criminal basis. State and local governments also have this discretion in many areas. Although there are important differences between criminal and civil laws, in the context of ex post facto laws it makes no sense to protect against criminal but not civil ex post facto laws. With either type, when the offending act was performed, it was not prohibited. Why should there be less protection against an ex post facto law if someone is sued on civil rather than criminal charges?

It might be thought that criminal laws can be distinguished from civil laws because violations of criminal laws, unlike civil laws, can be punished with physical incarceration. This reasoning fails, however, because incarceration as punishment for violation of civil law was prevalent in colonial times when the Constitution was drafted. For example, there were debtor prisons that required incarceration for nonpayment of debt. In fact, the existence of debtor prisons at the time of the Constitution shows that there was no hard and fast distinction between civil and criminal laws and is one more argument for applying the prohibition against ex post facto laws to civil laws.[2]

Due Process Concerns: Inconsistent Standards for Retrospective versus Prospective Legislation

While the Supreme Court has not issued a blanket prohibition against civil ex post facto laws, it has recognized that retroactive laws do present real problems of fairness. General Motors Corp. v. Evert Romein (1992) involved the constitutionality of a law passed by the Michigan legislature that required automobile companies to pay certain worker benefits retroactively. That law was passed to undo a Michigan Supreme Court decision (interpreting a previous Michigan worker's compensation statute) that said benefits were not to be paid retroactively. General Motors and Ford sued on the grounds that requiring new benefits to be paid retroactively violated the contracts clause. The Court held that the parties never explicitly bargained for the required benefits. Hence, the Court held that there was no violation of the contracts clause and that the ex post facto law that required new benefits to be paid retroactively was constitutional.

The General Motors case would have been easy if the prohibition on civil ex post facto laws was intact: the subject law was obviously retroactive and would have been prohibited. Instead, General Motors and Ford had to pay an additional $25 million in retroactive benefits. Justice Sandra Day O'Connor identified the unfairness of retroactive legislation (General Motors: 320):

Retroactive legislation presents problems of unfairness that are more serious than those posed by prospective legislation, because it can deprive citizens of legitimate expectations and upset settled transactions. For this reason, "the retroactive aspects of economic legislation must meet the test of due process''--a legitimate legislative purpose furthered by rational means.

Unfortunately, the due process standard of "a legitimate legislative purpose furthered by rational means'' is so easily satisfied that it renders any due process constraints against retroactive laws of very dubious use. Virtually any law passed by a majority can be said to satisfy a legitimate legislative purpose. Similarly, a test for "rational means'' is vague. An activist court can second-guess the legislature and make the rational means standard difficult to satisfy, while a court of judicial restraint can decide that any piece of legislation satisfies the rational means standard.

The whole purpose behind the prohibition on ex post facto laws, as the Federalist Papers note, is to give a regular course to the business of society and banish speculations on public measures. Allowing interest groups to lobby a legislature to pass ex post facto laws, and abiding by the results so long as they pass the basically standardless criteria of "a legitimate legislative purpose furthered by rational means,'' does not provide the regular course to business that the Founders intended when they included the constitutional prohibition on ex post facto laws.

It is interesting to contrast the due process standard with respect to vagueness with the due process standard regarding retrospective legislation. Regarding vagueness, the Court has held in cases such as Grayned v. City of Rockford (1972) that government regulation must be sufficiently clear so that ordinary people can understand what conduct is being prohibited. In Village of Hoffman Estates v. The Flip Side, Hoffman Estates, Inc. (1982: 498), the Court held that although "economic regulation is subject to a less strict vagueness test'' than criminal laws, vagueness analysis still applies. The Court gave two reasons for such a lesser standard. First, "because its subject matter is often more narrow, and because businesses can be expected to consult relevant legislation in advance of action'' (ibid.). Second, because "the regulated enterprise may have the ability to clarify the meaning of the regulation by its own inquiry, or by resort to an administrative process'' (ibid.: 455).

It should be clear that neither of these considerations supports the application of retroactive laws to economic legislation. First, business will not be able to consult relevant legislation in advance since the legislation is retroactive. Second, business will not be able to clarify the meaning of the regulation by resorting to administrative processes, because the retroactive legislation does not even exist at the time.

The contrast between what due process requires of retroactive and prospective legislation is stark. Due process requires that retrospective legislation meet the vague criteria of a "legitimate legislative purpose furthered by rational means.'' With respect to prospective legislation, on the other hand, due process requires that an ordinary citizen be able to know what is proscribed by a law. Barring clairvoyant abilities, the ordinary citizen will not know what behavior, including economic behavior, is proscribed by retroactive legislation. That is, retroactive legislation will certainly not be able to satisfy the due process standard required for prospective legislation. This seems an altogether unsatisfactory state of affairs. It makes no sense for due process to require that legislation be sufficiently clear so that an ordinary person must know what is proscribed in advance, and yet for due process to allow that same individual to become befuddled by retroactive laws that were impossible to know at the time of his or her actions.

United States v. Carlton: Modern Difficulties of Civil Ex Post Facto Laws

Recently, the Supreme Court unanimously held that retroactive tax laws were constitutional in United States v. Carlton (1994). While acting as an estate executor, Carlton relied upon a provision of the Tax Reform Act of 1986 that gave companies a favorable deduction from the proceeds of sale of a company's securities to its Employee Stock Ownership Plan. The government then proceeded to retroactively deny this favorable deduction by amending the 1986 law. The court of appeals in Carlton struck down the amended law as unconstitutional after finding that the estate had detrimentally relied upon the 1986 law and had no notice of the impending change. The court of appeals also cited the fact that retroactive income tax laws have been allowed but only for the year of enactment and the year prior to enactment (thus about a two-year limit). This decision was the first such invalidation of a retroactive tax law in almost half a century.

The Supreme Court, however, reversed the decision by the Court of Appeals for the Ninth Circuit and found that due process was not violated by retroactive application of the amended tax statute. According to the Court, the retroactive application of the tax amendment was consistent with the Fifth Amendment's due process clause because: (1) The purpose of the congressional amendment, which was to avoid a significant and unplanned federal revenue loss, was neither illegitimate nor arbitrary; (2) Congress acted promptly and established the retroactivity for only a short duration; (3) Carlton's reliance was insufficient to establish a constitutional violation; and (4) Carlton's lack of notice regarding the retroactive changes was not dispositive.

Although the vote was unanimous, Justice Antonin Scalia expressed misgivings about retroactivity in his concurring opinion, which Justice Clarence Thomas joined. Justice Scalia expressed reluctance to endorse the ineffective standard of a "legitimate legislative purpose furthered by rational means,'' although Scalia nevertheless concurred with the majority because of his distrust of the concept of "substantive due process.'' After castigating the "bait and switch'' policy represented by retroactive taxation, Justice Scalia still upheld retroactive taxation, although it is clear from his remarks regarding "bait and switch'' taxation that he did not feel comfortable ratifying the government's arbitrary conduct. Indeed, Scalia's opinion reads more like a dissent.

In his opinion, Justice Scalia labeled the term "substantive due process'' an oxymoron. Although he does not elaborate, his idea seems to be that any substantive right cannot by definition be protected by a formalistic, procedural concept such as "due process.'' Scalia skeptically notes that substantive due process has been used to invent rights such as the right to abortion in Roe v. Wade (1973) and the right to decide family living arrangements in Moore v. East Cleveland (1977) that he cannot find in the original constitution. Ultimately, Justice Scalia also rejects the expansion of substantive due process rights in the context of Carlton.

The legal brief presented to the Court for Carlton did not advance a frontal challenge to the constitutionality of civil ex post facto legislation. In fact, Calder v. Bull was not even mentioned in the brief. Rather, Carlton argued that the retroactive taxation violated substantive due process because it was "harsh and oppressive.''

Scalia could have used Calder v. Bull to legitimately strike down the retroactive taxation in Carlton, without succumbing to substantive due process. For example, Carlton's argument could have addressed the Constitution's explicit prohibition of ex post facto laws and included an analysis of the flawed logic of the Calder decision. The lawyers in Carlton were in the unenviable position of arguing with one hand tied behind their backs when they attempted to challenge the government's retroactive actions without referring to Calder.

In addition, the law in Carlton could have been overturned based on the argument that it violates procedural due process. The rather vague standard for substantive due process ("unduly harsh and oppressive'') is unlike the standard for procedural due process, which requires that an individual have notice of laws and an opportunity to be heard. The retroactive taxation could have been struck down on the grounds of procedural due process without invoking Scalia's concerns about the oxymoronic nature of substantive due process.

Stephen Munzer: Is Justice a Basis for Retroactive Legislation or Is It Duplicity?

There is one prominent author in the literature who explicitly attempts to provide a normative basis for retroactive legislation. It is to Stephen Munzer's credit that he does provide such an explicit justification, beyond the modern Court's rationale that a "legitimate legislative purpose furthered by rational means'' justifies such legislation.

According to Munzer (1982), the prohibition against retroactive laws relies heavily on the justice of the existing laws which cannot be retroactively changed. For instance, Munzer notes that laws allowing all Nazi's to escape punishment for killing Jews, or slaveholders to keep their slaves forever, should not be immune from retroactive legislation (ibid.: 434). Such examples lead Munzer to conclude that there should not be a total prohibition against retroactivity. Instead, Munzer argues that there should be a presumption against retroactivity that can be overruled when existing laws are deemed unjust.

Although Munzer initially provides examples that evoke general sympathy, he goes on to espouse the more problematic view that redistributional considerations should generally permit retroactive laws. Munzer states that his general argument does not presuppose any particular conception of justice. However, he expresses sympathy for a view of justice close to the view advocated by John Rawls (ibid.: 435-9). Rawls (1971) argues that the basic institutions of society should be organized to make the least advantaged individuals (or groups) in society as well-off as possible. He calls this the "Difference Principle.'' Rawls embraces an egalitarian framework in which potentially large gains of material well-being for the rest of society are held hostage to the incremental gains for those who are "worse off'' in society.

Munzer argues that cases involving retroactivity should be filtered through this redistributivist prism. If the poor are favored by retroactive laws, or if the rich are disadvantaged while society benefits, then a strong case can be made for allowing a retroactive law. In U.S. Trust Co. v. New Jersey (1977), in which bondholder rights were retroactively weakened, Munzer supports the Court's holding that retroactivity should have been allowed. U.S. Trust involved a bond covenant that barred the Port Authority from investing in mass transit projects that incurred deficits beyond a certain limit. The Court reasoned that the new projects the Port Authority was barred from undertaking would have helped society in general while prosperous bondholders would have been only marginally disadvantaged. With this logic, the Court held that retroactive legislative action breaking the bond covenant was permissible. According to Munzer, a worthwhile societal aim was furthered by the weakening of bondholder rights.

Munzer's argument for retroactivity contends that Rawls's redistributivist view allows for weaker adherence to the rule of law than Hayek's view. What Munzer overlooks is the element of publicity. Publicity is one of the major arguments for Rawls's two principles of justice and a constraint upon any acceptable outcome from a Kantian-contractarian framework in which rational actors knowingly choose basic principles by which they live.[3] The basic laws of society must be known to the public. It is impossible to square publicity in Rawls's theory with retroactivity where, by definition, citizens will not be able to know the basic laws as they exist at the time.

Another problem with arguing for retroactivity from egalitarian redistributional principles is the existence of the progressive income tax. The basic laws of a progressive income tax already reflect the redistributional impulses of the democratic majority. As long as those tax laws are prospective, individuals know that if they work in the future, they must give a percentage of their earnings to the government for egalitarian redistribution. Retroactive redistribution in addition to the income tax is problematic because society has basically set a limit to the amount of redistribution it considers fair via prospective progressive income tax laws. Any further redistribution forced by retroactive laws is basically deceitful as it cannot be passed in advance by a democratic majority contemplating a chosen level of redistribution.

It is therefore somewhat self-defeating to state a coherent, specific principle allowing retroactivity. For example, if bondholder covenants could be retroactively weakened or employers could be made to retroactively increase benefits, the redistributional benefits of retroactivity would disappear as bondholders and employers would take actions to adjust their positions prospectively. Given the existence of the prospective progressive income tax to accomplish redistribution, it is problematic to sacrifice the rule of law for the minor redistributional changes resulting from the necessarily surreptitious results of retroactivity.

Andrew Weiler: Criteria for Curtailing Retroactivity

Some scholars have noted the unsatisfactory state of affairs with respect to retroactive legislation and have proposed criteria to make it more difficult to sustain retroactive laws. In his 1993 article "Has Due Process Struck Out? The Judicial Rubberstamping of Retroactive Economic Laws,'' Andrew Weiler argues that the judiciary has given a rubberstamp to retroactive economic laws under the due process standard. Although Weiler does not directly challenge Calder, he argues that the Supreme Court has been too lax in allowing retroactivity in certain cases. He suggests that four criteria be used in analyzing retroactivity: notice, reliance, equity, and moderation of the burden. Weiler states that these factors should help reign in the excesses of retroactivity (Weiler 1993: 1126-31).

The first criteria is that of notice of the new law. Courts should find adequate notice only if the retroactivity period begins after the law's initial publication. This would include, for instance, the introduction of any bill into Congress that states that its effective date will be from its introduction rather than from its enactment.

This criteria of notice is of very little use in curtailing retroactivity because it does not enable individuals to plan their actions based on known rules. Competing and inconsistent bills are frequently introduced into Congress and an individual cannot realistically have notice as to which one will eventually be passed. Indeed, any new bill that is introduced will be in conflict with current law; thus an individual will not know whether to follow current law or one of any number of new bills.

Unfortunately, this rather weak criteria of notice is made even less useful in Weiler's framework. He states that even if there is inadequate notice, the retroactive law need not be automatically invalidated (ibid.: 1129) Rather, it is simply one factor to be balanced against the other three factors.

The second element of reliance recognizes that a law may be less offensive if a party has no prior reasonable expectations. With respect to changes in income tax, Weiler argues that there is no reasonable reliance against changes in income tax laws; while with respect to Carlton, there is reasonable reliance as the government had encouraged a specific type of conduct with the favorable tax deduction. Weiler is referring to the specific encouragement the Carlton estate relied upon in selling via an employee stock ownership plan. He discusses the court of appeal opinion that found this retroactive law unconstitutional and that was later reversed by the Supreme Court.

Detrimental reliance, however, is not an adequate substitute for a prohibition on ex post facto laws. First, neither legislative enactments nor court decisions should be allowed to subvert the certainty that a prohibition on retroactivity yields. It is not acceptable to allow Congress to do an end-run around the prohibition on retroactivity by incorporating into its laws a warning that certain laws permit retroactivity. This tactic would allow a legislative enactment to override a constitutional prohibition.

The second reason that detrimental reliance is not an adequate substitute for a prohibition on ex post facto laws is that it requires and encourages ignorance of the fact that civil ex post facto laws are still permitted as a result of Calder v. Bull. Weiler contrasts the specific conduct encouraged in Carlton with the Court's strong historical acceptance of retroactive income tax laws. Weiler suggests that Carlton reasonably relied upon existing laws governing tax deductions in a way that an individual could not reasonably rely upon existing income tax laws (ibid.: 1129).

The problem with Weiler's distinction is that it is only valid if Carlton did not know that civil ex post facto laws are generally permitted. If he were more legally sophisticated and knowledgeable, he would have known that even though Congress was encouraging a specific type of conduct, it was free to change its mind retroactively for up to a two-year period. (This is true at least with respect to court decisions about income tax laws.) A theory that relies upon ignorance of the law is unsatisfactory. Short of a prohibition on all ex post facto laws, or at least certain types of civil ex post facto laws, an individual would have to be ignorant to reasonably trust that civil laws would not be applied retroactively.

A rule that requires detrimental reliance, even after the courts have seen fit to allow retroactive civil laws generally, is illogical. Does it then make sense to require that an individual must have actually relied to his detriment on an existing civil law known to him? After all, no one knows all the civil or criminal laws at the time of action. An individual is not shielded from laws because of such ignorance. It may be more egregious if an individual actually relied on them, but reliance is hardly necessary. Thus, while actual detrimental reliance may be sufficient to outlaw retroactivity (assuming the courts would disallow it), it is hardly necessary in civil (or criminal) law.

The last two elements of Weiler's balancing test are equity and moderation of the burden. Under the equity analysis, "the rationality of the government's interest is examined and compared with the private party's conduct. This analysis provides courts with the flexibility requisite to an analysis of the inherently vague protections of the Due Process Clauses'' (ibid.: 1130-1). Closely related to the equity component is the extent to which the legislature moderated the retroactive burden. The government must demonstrate the extent to which the legislature attempted to limit the severity of the changes to the prior law and the burden imposed upon the individual, and show that the length of the retroactive burden is necessary to accomplish the legislative goal. Unfortunately, the elements of equity and moderation of the burden lead us back to the due process standard that is vague in how it has been applied by the Supreme Court to retroactive laws.

Weiler is well-motivated by concern that the due process standard has resulted in the rubberstamping of retroactive laws that are objectionable. His analysis unfortunately stops short of studying the seminal Calder case that landed us in the predicament he laments. His patchwork solution would merely replace the vagaries of the due process standard with his obscure standards for equity and moderation of burden (as well as the other two factors mentioned).[4]

The Impact of Ex Post Facto Laws on Property Rights

The tension between permitting civil ex post facto laws and prohibiting the taking of private property was evident in Calder. Indeed, we may well ask how there can be private property without some prohibition on ex post facto laws. One of the key marks of private property is that the property vests in an individual. How can property vest in an individual if the government is always free to change laws retroactively?

For instance, suppose Congress wants to change the tax law so that certain individuals owe taxes on money made three years ago. Or even more radically, suppose Congress required that all individuals who had more than $100 in the bank now have to pay 90 percent of such money in tax. Would such manifestly unjust ex post facto laws be ruled unconstitutional?

Surprisingly enough, the Court has sustained such manifestly unjust retroactive tax laws.[5] These pernicious results are traceable to Calder. It is particularly objectionable that the federal government be allowed to pass civil ex post facto laws when we consider that the Court's main reason for allowing civil ex post facto laws was that the contract clause and legal tender clause would have been unnecessary if civil ex post facto laws had been encompassed within the scope of the prohibition on ex post facto laws. But given that these more specific prohibitions are not stated as restrictions against the federal government (although they are against the states), an argument for allowing the federal government to pass civil ex post facto laws is extremely weak.[6]

Protect Property Rights by Prohibiting Ex Post Facto Laws

Two examples are useful in illustrating how prohibitions on ex post facto laws serve to protect property rights. The first example occurs in California state law. The traditional rule in California land use law, as described in Lippert v. Avco Community Developers, Inc. (1976), was that there was no vested right to develop one's property until an individual had a valid building permit and had performed substantial construction in good faith reliance upon the permit. California law was changed by the legislature in recognition of the fact that the building permit plays a relatively trivial part in the development process today, as compared to general plan amendments, zone changes, and environmental impact reports. The old law allowed a city to enact new zoning laws right up until the time that a building permit had been issued, which could be many years after a property owner applied for the subdivision permits that must be secured before a building permit could be issued. The new law, California's Government Code st66474.2, states that no new zoning laws (ex post facto laws, for example) can be applied after an individual has applied for a subdivision permit (a tentative tract map, in the parlance of California). By prohibiting the retroactive application of these zoning laws to completed applications to subdivide, the legislature expanded the rights of property owners.

The second example of the interplay between ex post facto laws and property rights occurs in Lucas v. South Carolina Coastal Council (1992). David Lucas paid almost one million dollars for two beachfront lots. When he bought the lots, he was allowed to build one house on each lot. The legislature then passed new laws prohibiting any building on the lots that were applied retroactively to his purchase.

The Supreme Court struck down the legislature's retroactive laws via the takings clause because Lucas was left with no viable use of his land and because he did not present any nuisance to his neighbors.[7] However, what if the new South Carolina law had allowed all owners the right to build only one house on any remaining lots they own? According to the Court, such an ex post facto law would not be counted as a taking as all use of the property has not been taken. Yet such a milder ex post facto law still would be objectionable, because it would deprive individuals of large portions and uses of their property, and it would deprive them of the ability to know in advance which laws will apply to their property.

Justice William Kennedy's concuring opinion spoke about laws that violate "investment-backed expectations'' and serve as the basis for a takings claim. He stated that the Supreme Court of South Carolina erred in not making a determination that the new zoning restrictions were in accord with the owner's reasonable expectations and therefore sufficient to support a severe restriction on specific parcels of property.

The problem with a theory that relies upon investment-backed "expectations'' is the requirement of determining which expectations are reasonable and which are not. The obvious definition of "reasonable expectations'' is that an owner has the right to reasonably expect to develop property according to the rules in effect when he purchases the property and not to have new laws retroactively applied to his purchase of the property. It is thus much more straightforward to forbid ex post facto civil laws and formulate the rule that expectations which rely upon the existing laws are those which will be protected.

Scope: Drawing the Line on Reliance

The unfairness of ex post facto laws in some cases will be very clear, as in General Motors or Lucas. Some cases, however, are not quite as clear. For example, where does one draw the line when many actions form a larger plan of action, and the earlier actions are completed while the later ones are not? When a new law is enacted, how far back in the chain does one look to determine ex post facto effect? For instance, Lucas would protect a property owner from new zoning and building laws that are different and applied retroactively to his purchase. But what about new depreciation laws that are applied retroactively to his building--even though he has not yet started on the building?[8]

As a factual matter, it seems clear that the purchase of land and construction of a building are acts in the past that are affected when new depreciation schedules are imposed and thus would be protected from new laws. Would an action such as a mere trip down to a local zoning board be protected, even before there was any other action accompanying it and creating further reliance upon the existing laws? Probably not. Will there be some intermediate cases in which it is difficult to say if the past action is part of an on-going course of action that is being affected by the new law?[9] Obviously, yes.

This issue is no different from criminal law, however, where a conspiracy to commit a crime may take place over a long period of time, with the earlier actions being committed well before the later ones. If the law has been strengthened by the time the later actions are committed, but not the earlier ones, we encounter the same issue of defining the extent to which the earlier actions must be connected to the later ones for there to be an ex post facto situation. For instance, if active planning for a conspiracy took place, there is an integrated course of action. On the other hand, if only an unrelated trip to a library took place to gather information (which was coincidentally later used in the crime), there is no such integrated course of action and therefore no ex post facto situation.

In Takings (1985), Richard Epstein forcefully argues for a concept with respect to property that applies equally to ex post facto laws: just because there are some difficult cases does not mean that the concept should be discarded when there are so many clear cases.

The last section of this paper will propose a comprehensive solution to the problems created by ex post facto laws, but in the absence of a thorough solution, we will consider two situations in which ex post facto laws seem most objectionable. The first is when there are no further actions to be undertaken as part of a course of action. This is a pure ex post facto situation where a course of action has been completed and then the existing laws are changed. An example of this would be if the 1992 tax laws were changed so that an individual owes more money on his 1990 taxes.

This situation seems more objectionable than if the laws are changed on future taxes, even though those changes may frustrate the expectations of those who relied on the earlier law. Thus, for instance, it seems more objectionable to be told to pay an additional sum for your 1990 taxes than it is to be told that depreciation laws will change in the future years, even though an individual may have purchased property based on an expectation that there would be no such changes applied to his past actions like the purchase of the land.

The difference between the two examples is in degree, not kind. Even in the first case of the individual with an already completed course of action who is told he must pay more taxes on his 1990 returns, there is still the issue of whether he reasonably relied upon the 1990 laws not being changed when he filed. This question of reasonable reliance exists with respect to past actions just as much as the question of reasonable expectations exists with respect to future actions. In fact, barring an effective prohibition against ex post facto laws, any reliance upon existing laws is misplaced and unreasonable. Although the example of a retroactive billing on 1990 taxes may seem farfetched, it is logically equivalent to what happened in General Motors, where the Court told the automobile companies that they retroactively owed money on already completed actions. Applying the amorphous standard of a "legitimate legislative purpose furthered by rational means,'' a court could no doubt find that retroactive laws on taxes satisfy the standard as well.

The second type of ex post facto law that is particularly objectionable is where the future actions that are affected are well-articulated. If they are well-defined, there is not a problem of whether there is any abuse of the system if an individual claims that his prior plans (extending into the future) were frustrated.

One way for the plans to be well-defined is through a contract. This is one attractive feature of the contracts clause, which Calder rightly recognized as containing a prohibition on a certain kind of ex post facto laws. Although the contracts clause prohibits ex post facto laws on actions that have not been completed, at least there is a contract that spells out what the future contemplated actions are. Hence, because we know what these actions are, via the contract, there is less potential for abuse (for example, abuse by someone trying to feign that they were wrongfully affected when they actually were not). A contract is not the only way to set forth the future course of action. Instead, an edict for the future can come directly from the government as can be seen by California Government Code st66474.2, California's recent move to prohibit retroactive zoning laws after an individual has applied for a subdivision permit.

A third type of ex post facto law is one that deals solely with public property--for example, laws governing speed limits on freeways, kinds of vehicles allowed on roads, and access to parks. It seems that in these areas there is less claim for enforcement of the prohibition against ex post facto laws. For instance, if the speed limit is decreased, it does not seem that car manufacturers would have a constitutional claim based upon the fact that they geared their high end production cars to a higher speed limit. The roads belong to the public and recognizing expectations that prohibit policy changes with regard to publicly owned land might obstruct needed improvements.

Although these three types of ex post facto laws are not exhaustive, they are an attempt to provide some further guidance beyond that provided by the current standard that allows retroactive legislation as long as it satisfies the standardless criteria of a "legitimate legislative purpose furthered by rational means.'' Courts and legislatures would do well to be more wary of ex post facto laws of the first two types and to be more liberal in the last area. As we shall see in the last section of the paper, however, a comprehensive solution to ex post facto laws will obviate the need for specific proposals.

A Proposal: Provide Just Compensation for Unfair Retroactivity

Something more than the rather lax standard of a "legitimate legislative purpose furthered by rational means'' is needed to protect due process rights. Weiler's proposal highlights the problem but does not suitably solve it.

One suitable proposal for handling retroactive legislation is a more flexible approach that would allow retroactive laws so long as individuals harmed by the new laws are compensated. It would make retroactive analysis similar to Fifth Amendment takings analysis. For example, a taking for the public good is permissible if there is compensation to the individuals whose property is taken. Thus, if the state wants to impose new environmental clean-up laws on property owners retroactively, it must pay just compensation to them.

Given the close connection between a prohibition on retroactive civil laws and the safeguarding of private property rights, it should not be surprising that enlightened retroactivity analysis mirrors takings analysis. The Fifth Amendment, which allows takings for public use provided compensation is paid, came after the absolute ban on retroactive laws contained in the original Constitution. Therefore, the Fifth Amendment can be viewed as implicitly amending the absolute ban on retroactivity to allow retroactivity if compensation is paid to harmed individuals.

This proposal makes it possible to respond to one of the chief arguments that proponents of retroactivity have given, namely, that retroactivity should be permitted because the state needs flexibility in running its affairs (e.g., Weiler 1993). Allowing retroactivity if compensation is paid offers the state the flexibility that the Fifth Amendment allows; the state merely needs to pay for it. This flexibility, moreover, is not invented by relying on the vagaries of "substantive due process.'' It flows from the initial wording of the Constitution banning ex post fact laws and an analysis of the subsequently passed Fifth Amendment that allows takings as long as individuals are compensated.

Conclusion

F.A. Hayek was correct to point out that, if coercion cannot be eliminated, its ill effects can be curtailed by letting individuals know in advance how coercive laws will affect them. In order for people to plan for such laws, they must be known and certain. Subsequently, ex post facto civil and criminal laws must be prohibited.

Due to the Supreme Court's error-ridden analysis in Calder v. Bull, the prohibition against ex post facto laws has been applied only to criminal laws. As an analysis of United States v. Carlton and significant policy considerations indicates, this prohibition deserves to be extended to civil laws. Retroactive laws should be permitted only if just compensation is paid to harmed individuals.

References

Arroyo Vista Partners v. County of Santa Barbara (1990) 732 F. Supp. 1046, 1052.

Calder v. Bull (1798) 3 U.S. (3 Dall.) 386.

California Civil Code st3354 (Deering 1994).

California Government Code st66474.2 (Deering 1994).

Crosskey, W. (1953) Politics and the Constitution in the History of the United States, Vol. 1. Chicago: University of Chicago Press.

Epstein, R. (1984) "Toward a Revitalization of Contract Clause.'' University of Chicago Law Review 51: 703-51.

Epstein, R. (1985) Takings: Private Property and the Power of Eminent Domain. Cambridge, Mass.: Harvard University Press.

Field, O. (1920) "Ex Post Facto in the Constitution.'' Michigan Law Review 20: 315-31.

General Motors Corp. v. Evert Romein (1992) 503 U.S. 320.

Graham, R. (1993) "The Constitution, the Legislature, and Unfair Surprise: Toward a Reliance-Based Approach to the Contract Clause.'' Michigan Law Review 92: 398-437.

Grayned v. City of Rockford (1972) 408 U.S. 104, 108.

Hayek, F.A. (1960) The Constitution of Liberty. Chicago: University of Chicago Press.

Hochman, C. (1960) "The Supreme Court and the Constitutionality of Retroactive Legislation.'' Harvard Law Review 73: 692-727.

Lippert v. Avco Community Developers, Inc. (1976) 60 Cal. App. 3d 775.

Lucas v. South Carolina Coastal Council (1992) 112 S.Ct. 2886.

Madison, J. (1864a) The Federalist, No. 44. New York: Charles Scribner's Sons.

Madison, J. (1864b) The Federalist, No. 84. New York: Charles Scribner's Sons.

Maryland Constitution.

Massachusetts Constitution.

Moore v. East Cleveland (1977) 431 U.S. 494.

Munzer, S.R. (1982) "A Theory of Retroactive Legislation.'' Texas Law Review 61: 425-80.

Ogden v. Saunders (1827) 25 U.S. 213.

Rawls, J. (1971) A Theory of Justice. Cambridge, Mass.: Harvard University Press.

Roe v. Wade (1973) 410 U.S. 113.

Satterlee v. Matthewson (1829) 2 Pet (U.S., 1st. ed.) 685.

Smead, E. (1936) "The Rule Against Retroactive Legislation: A Basic Principle of Jurisprudence.'' Minnesota Law Review 20: 775-97.

United States Trust Company v. New Jersey (1977) 431 U.S. 1.

United States v. Carlton (1994) 114 S.Ct. 2018.

United States v. Darusmont (1980) 449 U.S. 292.

United States v. Hemme (1985) 476 U.S. 558.

Untermeyer v. Anderson (1928) 276 U.S. 440.

Village of Hoffman Estates v. The Flip Side, Hoffman Estates, Inc. (1982) 455 U.S. 489, 498.

Weiler, A. (1993) Has Due Process Struck Out? The Judicial Rubberstamping of Retroactive Economic Laws.'' Duke Law Journal 42: 1069-1140.

Welch v. Henry (1938) 305 U.S. 134.

Steve Selinger is a land developer who has been the "beneficiary'' of certain ex post facto laws. He holds a doctorate in philosophy from Princeton University and a doctorate in economics and econometrics from the California Institute of Technology. He thanks Lawrence E. Smith, Bill Talbott, Suzanne Tracy, and Beth Vella for their helpful comments on earlier drafts.

Notes

[1] The dangers of ex post facto laws can be found in many contemporary contexts. On January 19, 1996, the Supreme Court agreed to hear the Justice Department's challenge of a federal appeals court ruling in U.S. v. Winstar that the government breached its promise to savings and loan institutions in 1989 when it revised accounting laws by unilaterally rescinding a 40-year goodwill amortization period. With approximately 100 similar cases pending against the government, the government could be liable for as much as $10 billion if the Court upholds the lower court's ruling in favor of the plaintiff, Glenfed, Inc. In this case, a retroactive change in accounting contributed to the failure of many savings and loan institutions. This is more than a breach of contract matter; it is a clear example of the disastrous effects of ex post facto laws, a point that the respondents may explore in their argument before the Supreme Court.

[2] Crosskey (1953: 349) indicates that the real motivation of the Calder Court to allow ex post facto civil laws was to grant bankruptcy relief for debtors, as one of the former Justices was in hiding and trying to avoid debtor's prison.

[3] A contractarian framework, which Immanuel Kant employed, assumes that ethical principles are those contracted or agreed to by moral agents acting impartially.

[4] Weiler's article appears about 33 years after Charles Hochman's "The Supreme Court and the Constitutionality of Retroactive Legislation,'' which appeared in the Harvard Law Review in 1960. In his article, Hochman (1960: 697) noted, "Since the great variety of cases in this field do not lend themselves to sweeping generalizations, it seems inappropriate to attempt to develop an ideal scheme for the court to follow in cases involving retroactive statutes, nor even to offer a formula for predicting the result in any given class of cases.'' The need for predictability in the law is not satisfied by the loose balancing tests proffered in place of a straightforward prohibition on retroactivity.

[5] Several Supreme Court cases have upheld retroactive taxes including United States v. Carlton (1994). In Welch v. Henry (1938), a divided Court ruled that a Wisconsin tax, which was passed in 1935 and made retroactive to 1933 income, was constitutional. More recently, in United States v. Darusmont (1980) and United States v. Hemme (1985), the Supreme Court overruled two district courts that had held that retroactive tax laws were unconstitutional. It should be noted that today's Russian Constitution explicitly prohibits retroactive tax laws.

[6] See, for example, Untermeyer v. Anderson (1928). The Court has sometimes declared retroactive estate and gift taxes unconstitutional on the due process grounds that there has been a voluntary action based upon reliance of no taxation. However, individuals also voluntarily work in reliance upon no (or the same) taxation and receive dividends (rather than sell assets) based upon no (or the same) taxation. Hence, this distinction seems groundless and is yet another example of how the amorphous due process standard is no substitute for the clear prohibition on ex post facto laws expressed in the Constitution.

[7] Technically, the Court remanded the case back to the South Carolina Supreme Court to either make a nearly impossible finding that Lucas presented a nuisance or to strike the law down.

[8] Sadly, this possibility is more than a hypothetical example. The 1986 tax law radically changed depreciation schedules to buildings that had been purchased under the old laws.

[9] Thus, even with an explicit prohibition of ex post facto laws, in some cases there would still be the issue of Justice Kennedy's concern for reasonable expectations. Without such a prohibition, however, no reliance on past laws can be said to be reasonable. A prohibition on ex post facto laws is thus a necessary but insufficient condition for reasonable expectations against changes in the law where an ongoing course of action is involved.

The Cato Journal is published in the spring/summer, fall, and winter by the Cato Institute, 1000 Massachusetts Ave., NW, Washington, D.C. 20001-5403. The Views expressed by the authors of the articles are their own and are not attributable to the editor, editorial board, or the Cato Institute. Printed copies of the Cato Journal may be ordered by calling 1-800-767-1241. Back issues are also available on the Cato Institute Web site: http://www.cato.org. Email comments or suggestions to subscriptions@cato.org.