94 Harv. L. Rev. 1574, 1576-1583 (1981)

... Allen's approach is flawed in that it only tests for the accuracy and rationality of judicial comment after finding that the issue affected is constitutionally essential to culpability.

Restricting his test in this way is problematic for two reasons. First, Allen fails to appreciate how limited and difficult the application of a substantive limitation theory will be, and therefore does not realize that his approach would virtually eliminate the requirement of rationality in instructing jurors. Second, rationality is a necessary, albeit not sufficient, condition for the constitutionality of every criminal presumption, whether or not the presumed fact is a constitutionally essential element of a crime.


Allen's analysis begins with In re Winship and the puzzle that it posed. Winship announced the constitutional requirement that a state prove every "element'' of a criminal offense beyond a reasonable doubt. That holding raised doubts about any use of affirmative defenses. If the statutory definition of a particular crime recognizes an affirmative defense, then it is possible to see the absence of that defense as an element that the state must prove beyond a reasonable doubt. Such a reading of Winship gained currency in Mullaney v. Wilbur. In Mullaney, the Court struck down a statute allowing the prosecution to rely on a presumption to prove malice, a statutorily prescribed element of homicide. The effect of the presumption was to impose on the defendant the burden of rebutting it (in order to mitigate the crime) by showing that the act was committed in the heat of passion.

Two years later, however, in Patterson v. New York, the Court adopted a narrow reading of Mullaney. The Court upheld a statute that shifted the burden to the defendant, not by a presumption as in Mullaney, but by a requirement that the defendant prove an affirmative defense. The Court took as its starting point the strict legislative definition of the offense: Only if the state expressly defined malice as an "element'' of the crime would Winship require the prosecution to prove it beyond a reasonable doubt, unaided by any presumptions. Apparently, the state would be permitted to undercut the force of Winship merely by shifting elements from the prosecution to the defense.

Abhorrence of Patterson's arid formalism stimulated a wave of commentary, including a fine article by Professors Jeffries and Stephan. Jeffries and Stephan(1)* rejected the purely formalistic approach that would have left the Winship issue up to state definition. However, they found equally untenable a reading of Mullaney that would require the state to prove beyond a reasonable doubt each and every issue affecting stigma or penalty. Such an approach would effectively eliminate all evidentiary devices, including affirmative defenses, by forcing the state to choose between two extremes. Either it must require the prosecution to prove an element beyond reasonable doubt, or it must eliminate it altogether from the definition of the crime. Jeffries and Stephan could find no substantive or procedural justification for this Hobson's choice. Indeed, putting states to such a choice might lead to harsher, less discriminating definitions of crimes. Further, it would be paradoxical to deny states the opportunity to use moderating measures (such as affirmative defenses and presumptions) to shift the burden on an element of an offense, but to allow them to eliminate the element altogether.

Attempting to cut between the two extreme solutions to the Winship puzzle, Jeffries and Stephan confronted the need for a constitutional jurisprudence defining the elements a state must prove to convict someone of a crime. Drawing on classic common law notions of actus reus and mens rea, as well as on the eighth amendment's prohibition of cruel and unusual punishment, they attempted to sketch out substantive constitutional limitations on the minimum definition of crime and on the maximum punishment that can be imposed for a particular offense.

Although Jeffries and Stephan presented their ideas of substantive limitation merely as directions for analysis, Allen seizes the theory, without developing it further, and makes it the keystone of his integrated test.... [S]ubstantive limitation embodies requirements of act, intent, and proportionality of punishment. It is difficult, however, to give content to these components. The "act'' requirement is intended to guard against the possibility that the state would punish mere intention or status, an important but obviously limited protection given the minimal nature of the act necessary to meet the requirement.(2)31 The "intent'' requirement is meant to ensure that a defendant's conduct is morally blameworthy, but its content too is both limited and obscure. Some crimes, for example, require no showing of intent at all.(3)32 As to proportionality, Jeffries and Stephan recognize that there is "no way to calculate exact relationships between wrong done and punishment earned, for the perception of good and bad and harm and blame flows from normative judgments that defy precise quantification.''(4)33 The Constitution would at most require "some rough sense of proportion in the assignment of sanctions.''(5)34 Because any finer honing of a proportionality standard would require articulation of a coherent theory of punishment, the authors in effect suggest that the test apply only when punishment is excessive in light of all possible justifications.


Allen's test demands accuracy and rationality of evidentiary devices only in cases running afoul of a substantive limitation theory. The problem that this poses is illustrated by Tot v. United States,(6)49 in which the defendant was convicted for violating a federal prohibition against possession of firearms by a felon. Tot's conviction was in part based on a statutory presumption authorizing the jury to infer from mere possession of a firearm that the firearm had been obtained in an interstate transaction. The Court invalidated the presumption, finding no "rational connection'' between possession and interstate transport. Despite this irrationality, Professor Allen's theory leads to the conclusion that the case was wrongly decided, unless he were to argue that movement of the gun in interstate commerce was an issue constitutionally essential to culpability (as opposed to federal jurisdiction) under his proportionality analysis.

Allen ... fails to recognize that rationality serves an independent constitutional value. The authority and integrity of courts rest to a significant degree on the extent to which society perceives them to be fair and rational arbiters of justice. It is one thing for the courts to tolerate a legislature's being ambiguous or arbitrary in defining a crime or establishing sentencing limits, and quite another to permit a legislature to insist on judicial arbitrariness in adjudicating criminal prosecutions. A legislatively established presumption that lacks a rational connection between predicate and conclusion makes the court administering it appear to be arbitrary. The very statement of the contrary proposition suggests Gilbert and Sullivan rules of procedure: "Jurors, you may presume A from proof of B, even though there is no rational connection between the two.'' One need not look further for the roots of the rational connection test. It does not depend on any theory of clear legislative statement or substantive due process, but rather on the notion that the legitimacy of judicial process requires that issues to be proved in court be proved rationally.

1. *Jeffries and Stephan, Defenses, Presumptions, and Burden of Proof in the Criminal Law, 88 Yale L.J. 1325 (1979).

2. 31. Consider, for example, the act requirement associated with the crime of conspiracy, see, e.g., Yates v. United States, 354 U.S. 298 (1957), the act of an alcoholic in being publicly drunk, see, e.g., Powell v. Texas, 392 U.S. 514 (1968), and the "act'' of nonfea[chsance, see Jones v. United States, 308 F.2d 307, 310 & nn. 8-11 (D.C. Cir. 1962).

3. 32. See, e.g., United States v. Park, 421 U.S. 658 (1975) (interstate shipment of contaminated food); United States v. Dotterweich, 320 U.S. 277 (1943) (interstate shipment of adulterated food and mislabeled drugs).

4. 33. Jeffries & Stephan, supra.

5. 34. Id. at 1377.

6. 49. 319 U.S. 463 (1943).

div1.gif (1531 bytes)
Home | Contents | Topical Index | Syllabi | Search | Contact Us | Professors' Pages
Cases | Problems | Rules | Statutes | Articles | Commentary