Was Shelley v. Kraemer Incorrectly Decided? Some New Answers

Mark D. Rosen of Chicago-Kent College of Law recently published an article titled Was Shelley v. Kraemer Incorrectly Decided? Some New Answers, 95 Cal. L. Rev. 451 (2007), which won the 2006 Outstanding Scholarly Paper Award from the Association of American Law Schools. Professor Rosen is a Norman & Edna Freehling Scholar and respected constitutional scholar who has testified to Congress on constitutional issues.

In his article, Professor Rosen explores the famous case of Shelley v. Kraemer and offers an alternative rationale for the holding. His arguments are radical and credible, and the advantages that flow from adopting his suggestions are significant.

I. A brief history of Shelley v. Kraemer

In Shelley v. Kraemer, 334 U.S. 1 (1948), the Shelleys purchased a home through a deed that had a restrictive covenant. A restrictive covenant is a private agreement among property owners that live near each other to restrict certain uses of their land. Like a zoning law, it prevents the owners from doing everything that they want with their land, but unlike a zoning law, which is a public law, a restrictive covenant is a private agreement–a contract that “runs with the land” and is enforced as a property right. The restrictive covenant on the property that the Shelleys purchased stated that the property could only be occupied by persons of the “Caucasian race.” Kraemer, a neighbor, sued in state court to prevent the Shelleys, who were African-American, from occupying the home and to force them to return title to the previous owners. The Supreme Court held that enforcement of the racially restrictive covenant was unconstitutionally discriminatory. The holding itself is consistent with the development of American law–courts and legislatures have increasingly limited the effects of invidious discrimination. The reasoning, however, is the black sheep of anti-discrimination law. A unanimous Court reasoned that if the state court enforced the racially restrictive covenant, then the enforcement would be “state action” in violation of the Equal Protection clause of the Fourteenth Amendment. The Fourteenth Amendment only applies to state action (“nor shall any State deprive person of . . . property, without due process of law.” (emphasis added)), and to decide the case under the Fourteenth Amendment, it was necessary for the Court to find some kind of state action related to the racially restrictive covenant. The case is unusual because it is extremely rare that a court considers the enforcement of a private agreement to be state action. Furthermore, by basing the case on the Fourteenth Amendment, the Court held that making a racially restrictive covenant is legal even though enforcing it is unconstitutional. (“We conclude, therefore, that the restrictive agreements standing alone cannot be regarded as a violation of any rights guaranteed to petitioners by the Fourteenth Amendment.”) This holding is odd because courts generally enforce legal contract provisions.

As Professor Rosen points out, this decision has caused considerable headaches for courts, academics, and students. He surveys prior attempts by academics to fit the reasoning of Shelley into the American legal structure, but concludes that none of them have been widely accepted by the legal community. Some commentators have tried to reconcile Shelley by arguing that since the state plays a significant role in nearly all private action, that the distinction between public action (e.g., judicial enforcement) and private action (e.g., the creation of an agreement) is so minor as to be unimportant. These academics argue that the Fourteenth Amendment’s state-action requirement is illusory because public and private actions are inseparabale. Another attempt to explain Shelley’s reasoning argued that the state had delegated the zoning power to private parties; therefore, the racially restrictive covenant was state action. A third attempt suggested that courts should balance the rights of the parties involved. If the liberty and property interests of the discriminating party, for example, outweigh the equal protection rights of the party discriminated against, then courts should not find state action. Finally, many courts have simply limited Shelley to discrimination, or even to only race discrimination.

For each of these prior approaches, Professor Rosen has some praise and some cogent criticisms. He argues, however, that it is not necessary for the legal community to completely disagree with these approaches, and instead argues that courts should adopt his approach because it is superior to the alternatives–including the reasoning in Shelley itself. His approach has two components–both are creative, unique, and bold. First, he suggests an alternate rationale for Shelley’s holding based on the Thirteenth Amendment instead of the Fourteenth Amendment. Second, and even more boldly, he suggests that the legal community completely discard the Fourteenth Amendment reasoning in Shelley.

II. The Thirteenth Amendment as an alternative rationale for Shelley

Professor Rosen convincingly argues that the holding in Shelley can be based on the Thirteenth Amendment. Section One of the Thirteenth Amendment states that “[n]either slavery nor involuntary servitude . . . shall exist within the United States.” Professor Rosen points out that using Section One as the rationale for Shelley is foreclosed because the Civil Rights Cases, 109 U.S. 3 (1883), narrowed Section One to abolishing slavery, and that it does not prevent private actors from discriminating based on race. Section Two of the Thirteenth Amendment states that “Congress shall have power to enforce this article by appropriate legislation.” The Court stated in the Civil Rights Cases that Congress may use its Section Two powers to outlaw the “badges” and “incidents” of slavery, including some types of invidious discrimination by private actors. Professor Rosen therefore argues that Congress has the power to pass a law that would render the racially restrictive covenant in Shelley illegal.

Professor Rosen suggests not one, but two, different laws grounded in Section Two that could apply to the facts in Shelley. The property aspects of the case are addressed by a predecessor to 42 U.S.C. section 1982, which stated, at the time of Shelley, that “all persons born in the United States and not subject to any foreign power . . . shall have the same right, in every State and Territory, as is enjoyed by white citizens thereof to . . .purchase . . . real and personal property.” Professor Rosen argues that the Court could have applied this law in Shelley and reached the same holding. The racially restrictive covenant was an attempt to prevent non-white citizens from purchasing real property and that clearly violates this statute.

The contract aspect is addressed by 42 U.S.C. section 1981, which states that “[a]ll persons within the jurisdiction of the United States shall have the same right in every State and Territory to make and enforce contracts . . . as is enjoyed by white citizens.” The Court could have applied this law in Shelley and reached the same holding because the racially restrictive covenant prevented non-whites from making a contract for the sale of the real property and that also clearly violates this statute. Applying either of these statutes to the facts in Shelley has the added advantage of holding the racially restrictive covenant illegal, whereas the Fourteenth Amendment rationale does not.

The biggest hurdle to using these statutes as an alternative rationale for Shelley is that the Shelley Court was aware of both of them but chose not to use the statutes. Professor Rosen makes a number of arguments that establish that the necessary legal interpretations of those statutes were not readily available when Shelley was decided in 1948. Section 1982, for example, was thought by the Court to have been enacted under Congress’s Fourteenth Amendment Section Five powers, but later cases established that it actually was enacted using Congress’s Thirteenth Amendment Section Two powers. The article easily clears this hurdle and Professor Rosen’s argument that Shelley can be understood as a Thirteenth Amendment case is persuasive.

III. Interpreting Shelley’s “attribution rationale”

Professor Rosen not only shows that a Thirteenth Amendment rationale is consistent with American jurisprudence, he also argues that the Fourteenth Amendment reasoning is so inconsistent with the rest of American case law that it should be discarded. He starts by defining the Fourteenth Amendment holding from Shelley.

The Shelley Court attributed enforcement of a racially restrictive covenant to the state in the following passage (Shelley at 21 (footnotes omitted)):

The Fourteenth Amendment declares, “that all persons, whether colored or white, shall stand equal before the laws of the States, and, in regard to the colored race, for whose protection the amendment was primarily designed, that no discrimination shall be made against them by law because of their color.” Strauder v. West Virginia, supra, 100 U.S. at 307. Only recently this Court has had occasion to declare that a state law which denied equal enjoyment of property rights to a designated class of citizens of specified race and ancestry, was not a legitimate exercise of the state’s police power but violated the guaranty of the equal protection of the laws. Oyama v. California, 1948, 332 U.S. 633. Nor may the discriminations imposed by the state courts in these cases be justified as proper exertions of state police power. Cf. Buchanan v. Warley, supra.

To make it easier to parse the text, I will use this paraphrase:

1. Strauder holding: the Fourteenth Amendment prevents discrimination by the law because of color (or race).
2. Oyama holding: a state law that discriminates on race (or color) is not a legitimate use of the police power.
3. Buchanan comparison: Nor is discrimination by a state court a legitimate use of the police power.

Professor Rosen dubs this reasoning the “attribution rationale” and offers two definitions.

1. “[C]ourts [can] only enforce contractual provisions that could have been enacted into general law.”
2. “Because the state could not have enacted the provision at issue in Shelley, it followed that court enforcement of the restrictive covenant also violated the guarantee of equal protection.”

Using the paraphrase above, Professor Rosen’s definition of the attribution rationale is “situations such as the Oyama holding inevitably lead to the Buchanan comparison.” Said differently, #3 must follow from #2. (The “must interpretation.”)

Although Professor Rosen does not precisely explain how he arrived at these definitions, his interpretation is a fair reading of Shelley, and there are logical reasons to interpret it this way. First, the Court used parallel language to describe both the Oyama holding and the Buchanan comparison. This implies a close relationship between the two concepts. Therefore, it is a fair assumption that the Court intended #3 always to follow from #2. Second, in the text, the Buchanan comparison literally does follow the Oyama holding, which implies the “must interpretation.”

Lastly, the most important statement in the quote is: “Nor may the discriminations imposed by the state courts in these cases be justified as proper exertions of state police power.” and this is supported by the comparison citation to Buchanan. The Buchanan case held that a segregation law, which had the effect of prohibiting the sale of specific property to African Americans, was not a valid exercise of the state’s police power. It is fair to assume that the Court meant that anytime the reasoning in Buchanan would invalidate a law, Shelley would invalidate an equivalent contract provision.

IV. Proof by contradiction

To prove that Shelley’s attribution rationale is invalid and should be discarded by the courts, Professor Rosen advances his argument with a proof by contradiction. Also called reductio ad absurdum, this type of argument disproves a premise by showing that the premise inevitably leads to an absurd, impossible, invalid, or contradictory result.

I should caution that Professor Rosen never calls his argument a proof by contradiction or a reductio ad absurdum. Certainly, lawyers use proofs by contradiction all the time, yet lawyers rarely call it that. The phrase “proof by contradiction” only appears 28 times in “All Law Reviews, Texts & Bar Journals (TP-ALL)” from Westlaw’s database, and most of those articles deal with economic arguments. A search on all federal and state cases (ALLCASES) yields 202 results, but a closer look at those cases makes it clear that the courts in those cases were talking about contradictory proof–evidence–and not a type of legal argument. The legal community has a natural affinity for Latin phrases, and that might explain why “reductio ad absurdum” is found more frequently in the database. A search on all articles yields 978 results, and it appears 981 times in all cases. Two thousand references out of the millions of articles and cases in the Westlaw database suggests that typical lawyers don’t use this phrase. The slippery slope argument is the little brother of proof by contradiction, and lawyers make slippery slope arguments all the time. It is odd that the legal community has not incorporated the efficiency and simplicity that comes with explicitly using a proof by contradiction. Nevertheless, Professor Rosen expertly employs a reductio ad absurdum argument, even if he doesn’t call it by that name.

One should also note that as a type of logical argument, a proof by contradiction might use syllogisms, but it doesn’t necessarily have to follow any syllogistic format.

With this type of proof, the ultimate conclusion is shown to be absurd because it contradicts something that is known to be true. It is acceptable to work through a series of logically true statements (called inferences) that ultimately lead to an inconsistent result. This example is from the Internet Encyclopedia of Philosophy:

1. Major premise: If A is true then B is true.
2. Minor premise: A is true.
3. Conclusion: Therefore, B is true.

Contradictory fact: B is false.

The conclusion is contradicted by the fact, so the major premise must be false. I use the term “premise” a little loosely; technically, a premise is a claim, but it might not be true. For the purposes of proof by contradiction, however, the minor premise is true or assumed to be true. This is necessary in a proof by contradiction. Because the major premise is the only part of the proof by contradiction that is not assumed to be true, when the proof leads to an absurd result, the only logical conclusion is that the major premise is false. If there were multiple premises in the proof, then the absurd result would call all of the premises into question. This is undesirable and similar to scientific experiments that have more than one independent variable–it is impossible to attribute the results to only one of the variables. Scientists avoid this problem by having only one variable in their experiments. Said differently, the scientists “hold constant” all the variables but one. Similarly, lawyers should have only one major premise in their proof by contradiction that is not known to be true.

V. Applying a proof by contradiction to Shelley’s attribution rationale strongly suggests that the Fourteenth Amendment reasoning should be discarded

In his article, Professor Rosen summarizes his argument to discard the Fourteenth Amendment rationale as (Rosen, at 501-02 (footnotes omitted)):

[T]he virtually uniform rejection of Shelley’s Fourteenth Amendment approach . . . is evidence that the Shelley Court’s rationale . . . does not accurately explain its holding. Recall that Shelley demanded state action, found it in the judicial enforcement of a private covenant, and made the state responsible for the covenant’s substantive content by means of the Court’s attribution rationale. If later courts had followed the attribution rationale, then judicial enforcement of contracts limiting speech would have triggered constitutional review, as would testamentary wills that condition inheritance on a child’s marriage to a person of a specified religious faith. But . . . these outcomes have not materialized.

This Article’s Thirteenth Amendment approach far better explains the post-Shelley case law.

Even though Professor Rosen does not refer to his argument as a reductio ad absurdum, I believe that this is a masterful application of a proof by contradiction. Recasting his argument as a proof by contradiction using the example where courts do enforce contracts limiting speech looks likes this:

1. Major premise from Shelley: a provision that would be unconstitutional as a law is unenforceable as a contract provision.
2. Minor premise: Provision X, a provision in a law limiting First Amendment free speech rights, would be unconstitutional.
3. Conclusion: Therefore, if Provision X is a provision in a contract, it is unenforceable by the courts.

Contradictory fact: Courts, however, do enforce contractual provisions limiting speech that would be unconstitutional if enacted as a law. This example comes from United Egg Producers v. Standard Brands, 44 F.3d 940 (11th Cir. 1995) (holding that a settlement agreement between two companies where the parties agreed not to create advertisements that “disparage the other party’s product” is a restriction of “their First Amendment rights on commercial speech,” specifically refusing to apply Shelley, and further holding that judicial enforcement of the settlement agreement was not state action). The fact that courts enforce these contacts directly contradicts the conclusion; therefore, the major premise must be false.

The example using wills looks as follows:

1. Major premise from Shelley: a provision that would be unconstitutional as a law is unenforceable as a contract provision.
2. Minor premise: Provision X, a provision in a law that conditions inheritance by an heir on that heir marrying someone of a particular faith would violate the Establishment Clause of the Constitution.
3. Conclusion: Therefore, if Provision X is a provision in a contract, it is unenforceable by the courts.

Contradictory fact: Courts, however, do enforce provisions in wills like this even though it would be unconstitutional if enacted as a law. This example is based on two cases, Shapira v. Union Nat’l Bank, 315 N.E.2d 825, 827-28 (Ohio Ct. Com. Pl. 1974) (finding that the testator conditioned inheritance on the heir marrying a “Jewish girl,” holding that it is “clear” that the heir’s “right to marry is constitutionally protected from restrictive state legislative action,” and holding that Shelley did not apply because the court was not asked to enjoin a marriage) and Gordon v. Gordon, 124 N.E.2d 228, 235 (Sup. Jud. Ct. Mass. 1955) (holding that a will conditioning inheritance on the heir marrying someone “born in the Hebrew faith” was not a violation of the First or Fourteenth Amendments because the will did not condition the inheritance on anyone’s religious belief, implying that the Amendments might apply if the will did condition the inheritance on someone’s faith, and holding that the district court can enforce the provisions of the will). Again, the fact contradicts the conclusion; therefore, the major premise must be false.

Professor Rosen uses these and other examples to show that the Fourteenth Amendment attribution rationale has been rejected by the courts. Since courts have rejected the reasoning, he argues that the legal community should discard it as the basis for the holding in Shelley, and adopt the Thirteenth Amendment rationale.

VI. Advantages to adopting the Thirteenth Amendment reasoning and discarding the Fourteenth Amendment attribution rationale

Besides his two major logical arguments, Professor Rosen makes seven normative arguments in favor of adopting the Thirteenth Amendment reasoning and discarding the Fourteenth Amendment attribution rationale.

1. The Thirteenth Amendment reasoning is more compelling.
2. Discarding the attribution rationale relieves stress on the distinction between public action and private action.
3. The article distinguishes #2 from “state action doctrine” and points out that discarding the Fourteenth Amendment reasoning will make it easier to formulate a clear state action doctrine.
4. The Thirteenth Amendment reasoning would make racially restrictive covenants illegal–not just unenforceable.
5. As shown with the proof by contradiction, rejecting the Fourteenth Amendment attribution rationale is more consistent with existing case law.
6. Adopting the Thirteenth Amendment reasoning would give new life to the amendment.
7. Grounding the case in Congress’s Thirteenth Amendment Section Two powers instead of Fourteenth Amendment constitutional prohibitions means that Congress and the President become more involved in shaping this area of law. The increased involvement of these two institutions brings the advantages of the different institutional competences to the table (such as increased fact-finding abilities), which would have a positive effect on the development of anti-discrimination law.

VII. The “might interpretation”: a different interpretation of Shelley’s attribution rationale

Professor Rosen’s proof by contradiction of Shelley’s attribution rationale is a strong argument for discarding the Fourteenth Amendment reasoning of the case. It relies on the “must interpretation” from Part III, above: if a contractual provision could not be enacted as a law, then it inevitably follows that courts cannot enforce the contractual provision. As shown in Part III, there are three reasons why the “must interpretation” is a reasonable reading of Shelley. It is also reasonable, however, to interpret the text to mean: if a contractual provision could not be enacted as a law, then it might follow that courts cannot enforce the contractual provision. (The “might interpretation.”)

First, the American legal system avoids using absolutes. At best, American law is full of exceptions; at worst, American law is languidly equivocal. Even a law as clear and obvious as “do not intentionally kill” has a long list of exceptions and mitigations. It is rare that legal statements are as strong as “If X, then Y inevitably follows.” Although the Shelley text uses parallel language to describe both the Oyama holding and the Buchanan comparison, the Court does not explicitly say whether it must follow or if it might follow. Said differently, the Court did not explicitly define the relationship between the two statements. Because American law avoids absolutes, it might be difficult to read Shelley as stating the “must interpretation.” This factor suggests that the “might interpretation” is more likely.

Second, the order of the statements does not inevitably lead to the “must interpretation.” The Court discussed the Oyama holding and then discussed the Buchanan comparison, and it is tempting to infer from the order of presentation that #1 inevitably leads to #2. Logically, this is not necessarily true. Rain usually precedes lightning, but rain does not inevitably lead to lightning. (Furthermore, lightning can sometimes happen without rain.) Similarly, the order of presentation does not necessarily mean that the “must interpretation” is the only interpretation of the text. This factor also suggests that the “might interpretation” is probable.

Third, the presumption that the Court chooses its words carefully suggests adoption of the “might interpretation.” When construing statutes and interpreting prior opinions, the Supreme Court presumes the drafters chose their words carefully. Think of Chisom v. Roemer, 501 U.S. 380 (1991), and the close reading that the majority inflicts on the Voting Rights Act. Then think of Justice Scalia’s dissenting opinion with an equally close reading where he practically throws Webster’s Second New International Dictionary across the chambers. Decisions can turn on the meaning of one word as shown by the epic battle of the dictionaries in Muscarello v. U.S., 524 U.S. 125 (1998).

Even introductory signals for citations can influence how the Supreme Court decides a case. According to The Bluebook: A Uniform System of Citations, “[s]ignals organize authorities and show how those authorities support or relate to a proposition given in the text.” (The Bluebook is not available online, but Wikipedia explains the concepts reasonably well.) When a citation is introduced without a signal, then the authority “directly states the proposition.” The citation signal “see” is used when the “authority clearly supports the proposition . . . [and] the proposition is not directly stated by the cited authority but obviously follows from it; there is an inferential step between the authority cited and the proposition it supports.” (Emphasis added.) The signal “cf.” (Latin for “compare”) means the “[c]ited authority supports a proposition different from the main proposition but sufficiently analogous to lend support.” (Emphasis added.)

The Supreme Court obviously takes these different signals seriously. In Lambrix v. Singletary, 520 U.S. 518 (1997), the Supreme Court had to decide if the rule announced in Espinoza v. Florida, 505 U.S. 1079 (1992), was dictated by the holding in Baldwin v. Alabama, 472 U.S. 372 (1985), or if Espinoza announced a “new rule.” Writing for the majority, Justice Scalia considered the introductory signal used by the Espinosa Court an important part of the meaning (Lambrix at 528-29 (footnote omitted)):

In our view, Espinosa was not dictated by precedent, but announced a new rule. . . . It is significant that Espinosa itself did not purport to rely upon any controlling precedent. The opinion cited only a single case, Baldwin v. Alabama, 472 U.S. 372, 382 (1985), in support of its central conclusion. . . . Espinosa, 505 U.S., at 1082. And it introduced that lone citation with a “cf.”–an introductory signal which shows authority that supports the point in dictum or by analogy, not one that “controls” or “dictates” the result.

By using the cf. signal, the Espinosa Court was saying that the holding in Espinosa was not dictated by the reasoning in Baldwin. Scalia’s implication is that if the Espinosa Court had used a stronger signal, such as “see,” then the holding in Espinosa may have been dictated by the reasoning in Baldwin. Consider that the “see” signal means that the stated proposition “obviously flows from” the reasoning in the authority. If the Espinosa Court had cited to Baldwin with a “see” signal then the Court would have been saying that the holding in Espinosa obviously flows from the reasoning in Baldwin.

Like the Lambrix, Chisom, and Muscarello Courts, the Shelley Court closely read prior case law to support its reasoning. To explain its holding that actions by state courts are certainly “state action” under the Fourteenth Amendment, for example, the Court painstakingly examined the language of dozens of prior cases. When parsing the Civil Rights Cases specifically, the Court stated, “Language to like effect is employed no less than eighteen times during the course of that opinion.” In other words, to support its reading of the Civil Rights Cases, the Court counted how many times a particular phrase was used. Just as the Shelley Court closely read prior cases, the Court would probably expect the Shelley opinion to be read closely, and the Court therefore likely chose its words–and signals–carefully.

In Part III, above, the Buchanan comparison was interpreted as supporting the “must interpretation.” Based on the signals, however, there might be a better interpretation. Shelley stated: “Nor may the discriminations imposed by the state courts in these cases be justified as proper exertions of state police power. Cf. Buchanan v. Warley, supra.” Applying the reasoning in Lambrix, and the definitions from The Bluebook, the only conclusion is that the holding in Shelley does not inevitably flow from the reasoning in Buchanan. Said differently, Buchanan might sometimes lead to Shelley, but it is not inevitable. If the Shelley Court wanted to say that the Buchanan reasoning inevitably lead to the Shelley holding, then it would have used the “see” signal–not the weaker cf. signal.

To turn an old legal phrase around, the logic in Buchanan is sufficient to support the holding in Shelley, but it does not necessarily require the holding in Shelley.

These three reasons mean that it is fair to read Shelley as stating the “might interpretation.” Specifically, the attribution rationale says that if a contractual provision could not be enacted as a law, then it might follow that courts cannot enforce the contractual provision.

If one adopts the “might interpretation,” then the proof by contradiction doesn’t work. Using the free speech example, the proof would look like this:

1. Major premise from Shelley: a provision that would be unconstitutional as a law might be unenforceable as a contract provision.
2. Minor premise: Provision X, a provision in a law limiting First Amendment free speech rights, would be unconstitutional.
3. Conclusion: Therefore, a contractual provision with the same limitation might be unenforceable by the courts.

Fact: Courts do enforce contractual provisions limiting speech that would be unconstitutional if enacted as a law.

In this case, the fact does not contradict the conclusion; therefore, the major premise might be true. Notably, it also does not prove that the major premise is true.

Professor Rosen’s proof by contradiction requires the “must interpretation” of Shelley’s attribution rationale. If one adopts the “might interpretation,” however, the proof by contradiction does not prove or disprove Shelley’s attribution rationale.

Conclusion and coda

Professor Rosen makes a compelling argument for adopting a Thirteenth Amendment reasoning for Shelley especially because there are a number of advantages by doing so. Furthermore, he makes a strong argument using a proof by contradiction that the legal community should discard the Fourteenth Amendment rationale. Even if one adopts the “might interpretation” above, Professor Rosen makes many compelling normative arguments in favor of discarding the Fourteenth Amendment rationale. His suggestions are compelling and the legal community should carefully consider them.

Finally, if the above is not enough to entice you to read his article, Professor Rosen has an extra bonus that certainly makes it worth your time. He introduces a concept he calls “constitutional preemption” where he translates the specifics of the dormant Commerce Clause doctrine into a general legal principle. It is a fascinating and persuasive concept, and one hopes that he will expound on it in a future article.

My thanks go to Professor Henry H. Perritt, Jr., Professor Rosen, and Professor Carolyn Shapiro for their help.