Burden of Proof

Commentators have suggested different approaches for deciding which issues should be governed by the reasonable doubt rule. At one extreme, the rule might apply to every issue, without exception, governing the proof of every fact that the criminal law makes relevant to a criminal conviction. At the other extreme, the rule might apply only to those issues for which the legislature has made no explicit exception. The Supreme Court has plainly rejected both extreme positions in Mullaney and Patterson. Under these cases some exceptions are permitted, but there are constitutional limits on the legislature's power to make exceptions. Although the Court has not specified the criteria for permissible exceptions, commentators have suggested several.

First, and least controversial, is the view that the Constitution permits an exception for issues in a criminal case that do not directly relate to guilt or innocence. In the course of a criminal prosecution, it may be necessary to decide whether the case is properly before the court, whether particular items may be admitted into evidence, or whether the defendant is mentally competent to stand trial. These decisions may well determine whether it is possible as a practical matter to convict the defendant, but they do not determine whether the defendant is in fact guilty. For that reason, the Constitution does not impose the reasonable doubt rule on such determinations, although the rule may nonetheless be required as a matter of state law.

A second, more controversial proposal is an exception for issues that present special problems of proof. It is suggested that the defendant should bear the burden of proof on an issue if the defendant has better access than the prosecution to the evidence. The rationale is that a defendant with control over the relevant evidence has a great incentive to withhold the evidence, mislead the jury, and prevail because of the prosecution's inability to meet its burden of proof. This strategy could be prevented by a rule shifting the burden of proof to the defendant. On this theory, the burden of proof might be assigned to the defendant on the issue of insanity or of intent. The problem with shifting the burden to the defendant for this reason is that it accomplishes too much. It not only elicits evidence from the defendant, but it also continues to tilt the scales against the defendant even after the evidence has been produced. A better solution to the problem of access to evidence would shift to the defendant the burden of coming forward with enough evidence to raise the issue, and then leave with the government the ultimate burden of proof after all the evidence is in.

A third proposed criterion for identifying exceptions to the reasonable doubt rule has become the center of a major debate. This controversy raises basic questions about the relationship between substantive law and procedure, as well as about the relationship between state legislatures and the federal Constitution on matters of criminal law. Some commentators have argued that the reasonable doubt rule should not apply to any fact that the legislature could constitutionally have omitted from its substantive criminal law. They argue that if the legislature has the constitutional power to make a fact irrelevant to guilt, then it must also have the power to choose its own rules for proving that fact. Put differently, if the legislature has created a gratuitous defense, then that issue is exempt from the requirement that the government prove its case beyond a reasonable doubt.

Other commentators argue that legislative power to eliminate a defense does not entail the power to shift to the defendant the burden of proof. They maintain that both the practical and the symbolic functions of the reasonable doubt rule apply with full force where a gratuitous defense is concerned.

The controversy is set forth in a pair of articles by Barbara Underwood and by John Jeffries and Paul Stephan. Jeffries and Stephan argue that it is both illogical and unwise to impose strict procedural requirements on the proof of a gratuitous defense. It is illogical, they say, because only if the Constitution requires the state to prove a particular fact as a prerequisite to conviction does the Constitution also require the state to prove that fact beyond a reasonable doubt. It is unwise, they argue, because legislatures have often been willing to enact new defenses to crime only in conjunction with rules that shift the burden of proof to the defendant. To prohibit such compromises, they contend, would stifle criminal law reform.

Underwood argues, by contrast, that the power to eliminate an issue from the criminal law does not entail the power to alter the rules of proof for that issue. In her view, the Constitution allocates to the states very broad power to define the substantive criminal law, but it imposes rigorous procedural requirements on the process of proving whatever facts the state has made criminal. Thus, the Constitution does not permit a state to adopt a controversial defense, and then limit it by shifting the burden of proof to the defendant. Instead, the state must resolve controversies over criminal law policy by making adjustments and compromises in the content of the substantive criminal law.

For example, a state legislature might be divided over a proposal to exempt from the narcotics law those who possess narcotics solely for personal use. A procedural compromise would be to adopt the defense in full, and to limit it by shifting the burden of proof to the defendant. A substantive compromise would be to adopt the defense in part, exempting only those who possess narcotics for personal use in specified small quantities, or in the privacy of the home. If the Constitution prohibits the procedural compromise, then the legislature must adopt or reject the defense or find a substantive compromise.

It is, of course, difficult to determine whether in fact criminal law reform would be stifled if the burden-shifting device were prohibited. Legislatures might instead adopt reforms without burden-shifting, or they might find satisfactory substantive compromises. Moreover, the argument that burden-shifting is necessary for legislative reform does not require an exception from the reasonable doubt rule for all gratuitous defenses. It requires an exception only for those new gratuitous defenses that result from legislative compromise.

Linking the reasonable doubt rule to only those facts that the Court says are constitutionally required to constitute a crime is equally questionable. Commentators have long urged the Court to develop a body of constitutional criminal law, but the Court has been highly reluctant to do so. In Montana v. Egelhoff, 518 U.S. 37 (1996), for instance, the Court struggled with the issue of whether the state could help meet its burden of proving that the defendant had intentionally killed the victim by barring him from showing that he was intoxicated. A bare majority of the Court held that the state could bar the intoxication defense, but it was badly split in its reasoning, and the case's difficulties suggest that the Court rarely will plunge into the quagmire of reviewing state substantive criminal law. Given the paucity of constitutional criminal law, if the rule applies only to constitutionally necessary facts, then it may have almost no application at all (Sundby, 1989a).

Moreover, the gratuitous character of a defense is not by itself sufficient to exempt that defense from other constitutional requirements of fair procedure. Even a defendant raising a gratuitous defense has the right to have that issue determined at a trial by jury, with counsel and confrontation of adverse witnesses. If an issue is exempt from the requirement of proof beyond a reasonable doubt, the reason must not be solely that the defense is gratuitous, but that for some reason it is less important to protect the defendant against error.