Attempt

To indicate the scope of liability for attempts in modern American law, the following will be considered: (1) the required state of mind, or mens rea; (2) the required acts, or actus reus; (3) liability when the offender desists before completing the intended crime (the problem of "abandonment"); (4) liability when the accused could not possibly carry out the intended crime (the problem of "impossibility"); and (5) the severity of punishment for attempt (the problem of "grading").

The mens rea. A criminal attempt is traditionally defined as an intent to perform an act or to bring about a result that would constitute a crime, together with some substantial steps taken in furtherance of the intent. In accordance with this definition, it is apparent that the state of mind, or mens rea, required is the actual intent or purpose to achieve the proscribed result; mere recklessness or negligence will not suffice.

The usual requirements of intention or purpose can appear anomalous when the many situations are considered in which the completed crime may be committed by recklessness, negligence, or even on a strict-liability basis. Suppose, for example, that a construction worker dynamites a hillside, with no intent to kill anyone, but with a reckless disregard for the lives of people residing nearby. If one of those people is killed by the explosion, the worker will be guilty of murder; recklessness is sufficient for liability. However, if the person injured by the explosion eventually recovers, the worker would not be guilty of murder and could not even be convicted of attempted murder because he was merely reckless and did not intend to kill.

How can this gap in attempts liability be explained? If in the event of death, the conduct should be punished as murder, then why does the identical behavior not remain a proper subject of penal sanctions when the victim luckily survives? One answer is definitional. An attempt, by the very meaning of the word, implies that the actor was trying to achieve the forbidden result, and this simply cannot be said of the construction worker. This view does not leave us with a very satisfying reason for not punishing the conduct, but rather focuses on the inelegance of referring to the conduct as an "attempt."

Some legislatures have relaxed to a limited degree the requirement of purpose or intention: one approach has been to create a specific offense of reckless endangerment, so that such conduct need not be prosecuted as an attempt (Pa. Cons. Stat. Ann. § 2705 (1983)).

The actus reus. The courts hold that certain preliminary activities, designated "mere preparation," are not punishable even when accompanied by the requisite intent. Attempt liability attaches only when the defendant goes beyond mere preparation and begins to carry out the planned crime.

How can one determine the location of this line dividing mere preparation from the punishable attempt? Suppose, for example, that a defendant announces his desire to blow up the office of a former employer, collects a supply of matches, old newspaper, and kerosene, buys dynamite and a long fuse, places the dynamite and other material in the building, and finally lights the fuse. At what point in this sequence of events has the defendant committed a punishable attempt? Cases confronting such questions have invoked a considerable variety of analytic devices and have come to widely divergent results. The most important approaches are those requiring either commission of the last necessary act, commission of an act proximate to the result, or commission of an act that unequivocally confirms the actor's intent. The Model Penal Code's approach combines elements of these three. After discussing these approaches, this section considers one other actus reus problem, the possibility of punishing "attempts to attempt."

The last-act test. Under the last-act test, suggested in Regina v. Eagleton, 6 E. Cox, Crim. Cas. 559 (C.C.A.) (London, 1855), the disgruntled employee in the example above would be guilty of attempt only after lighting the fuse. At that point, although the attempt may still miscarry, the actor has done everything that appears necessary to carry it through to completion. The last-act test is designed not only to ensure that the defendant's intent is serious, but also to provide an incentive for him to desist by enabling him to avoid liability right up to the last possible moment. The last-act test seems much too strict, however, in terms of the "early intervention" function of attempt law. A defendant who follows a victim, draws a gun, and takes careful aim could not be charged with attempt, because he had yet to pull the trigger. For these reasons, no contemporary court would insist strictly upon commission of the last necessary act (Model Penal Code, 1960, commentary on § 5.01). Attempt liability attaches at an earlier point, and the needed incentive to desist is provided by a separate defense of "abandonment," discussed below.

The proximity test. To avoid the practical difficulties of the last-act test, many courts apply a "proximity" test requiring only that a defendant's preparatory actions come rather close to completion of the intended crime. But how close is close enough? Two examples will indicate the difficulties of the proximity test. In Commonwealth v. Peaslee, 177 Mass. 267, 59 N.E. 55 (1901), the defendant arranged combustible material in a building and left. Later, intending to set off the blaze, he drove within a quarter mile of the building and then decided to turn back. Writing for the court, Justice Oliver Wendell Holmes suggested that this might be a punishable attempt. In People v. Rizzo, 246 N.Y. 334, 158 N.E. 888 (1927), the defendants spent considerable time driving around the streets of New York searching for a payroll clerk whom they intended to rob. The police arrested them before they could find the clerk, but the New York Court of Appeals held that this was not a punishable attempt. In such cases, courts may be thinking of proximity primarily in a physical or spatial sense; in Peaslee the defendant had driven most of the way to the building, whereas in Rizzo the payroll clerk was never located at all and the defendants seemingly never came "close" to actually putting their plan into action. However, this sort of spatial proximity is not only hard to specify, but totally unrelated to the purposes of attempt law. Preliminary acts should become punishable when they establish that the intent is likely to be put into action, that the individual is sufficiently dangerous to require restraint, and that there is a dangerous probability of success requiring deterrence and early police intervention. From all of these perspectives the case for punishment is at least as strong in Rizzo; indeed, the defendant in Peaslee was, if anything, less deserving of punishment because he apparently chose voluntarily to abandon his plan. Some courts have attempted to adapt the proximity test more satisfactorily to the purposes of attempt law by focusing on whether the acts involve a dangerous proximity to success or demonstrate that the actor was unlikely to desist, but these approaches also prove difficult to apply with objectivity and consistency.

The equivocality approach. Reluctance to punish "mere preparation" is based in part on concern that very preliminary acts may not confirm that the defendant seriously plans to put his intent into action. Accordingly, some authorities have suggested that to be punishable, a preliminary act must be "of such a nature as to be in itself sufficient evidence of the criminal intent with which it is done. A criminal attempt is an act which shows criminal intent on the face of it" (Rex v. Barker, [1924] N.Z.L.R. 865, 874 (C.A.)). American cases sometimes appear to speak approvingly of this requirement that the acts unequivocally confirm the criminal intent, and this approach does in theory appear consistent with the purposes of attempt law. Nevertheless, the equivocality approach, if applied literally, would often prove even stricter than the last-act approach. A defendant might approach a haystack, fill his pipe, light a match, light the pipe, and perhaps even toss the match on the haystack. The acts alone are not wholly unequivocal, but it is hard to imagine a court holding that regardless of any other evidence of intent, the acts themselves do not go far enough (Williams, p. 630).

The Model Penal Code approach. The Model Penal Code borrows from the concepts of proximity and equivocality but treats both in rather flexible fashion. Less suspicious of confessions and other direct evidence of intent, the Code relaxes the traditional insistence on very substantial preparation. Under the Code, an attempt must include "an act or omission constituting a substantial step in a course of conduct planned to culminate in . . . commission of the crime" (1962, § 5.01(1)(c)). The substantial-step requirement reflects proximity notions, but shifts the emphasis from the significance of the acts still required to the significance of what the defendant has already done. The Code also specifies that an act cannot be deemed a "substantial step" "unless it is strongly corroborative of the actor's criminal purpose" (§ 5.01(2)). The Code thus incorporates the concerns underlying the equivocality test, without being burdened by the impractical rigidity of that approach.

Attempts to attempt. Many substantive crimes are in effect attempts to commit some other offense. Assault, for example, is essentially an attempt to commit a battery; burglary (breaking and entering a structure with intent to commit a felony therein) is essentially an attempt to commit some other felony. Sometimes a defendant is charged with attempted assault or attempted burglary will argue that the alleged conduct should not be punishable because it amounts to no more than an attempt to attempt. Such arguments may suppose the conceptual impossibility of such an offense, or they may reflect the view that conduct not amounting to an attempt is necessarily "mere preparation." Neither position is plausible. Concerns about imposing attempt liability at an excessively early point need to be faced, but in principle there is no reason why preparations to commit burglary, for example, might not pass the realm of mere preparation, even though the burglary itself was not successfully perpetrated. Consider the case of a masked man caught in the act of picking the lock of an apartment door. In such a case, a charge of attempted burglary is clearly justified, and the courts so hold (Model Penal Code, 1960, commentary on § 5.01).

Abandonment. Once the defendant's conduct has moved from "mere preparation" into the realm of a punishable attempt, can he nevertheless avoid liability if he has a genuine change of heart and decides to abandon his plan? Many cases appear to give a negative answer to this question. Just as a defendant who has stolen property cannot avoid liability by making restitution, the courts often say that once the defendant's attempt goes far enough to be punishable, a crime has been committed and subsequent actions cannot change that fact, although they may have a bearing on the appropriate sentence (Perkins, pp. 319, 354).

Whatever the logic of this view, one of its consequences is to reinforce traditional objections to imposing liability at relatively early stages of preparatory conduct. In the absence of an abandonment defense, early liability eliminates a significant incentive to desist and appears unfair to the defendant who has had a genuine change of heart, as in Peaslee. Such concerns generate pressure to reject early liability even when there is no hint of possible abandonment by the defendants in the case actually at hand, as in Rizzo. In short, in the absence of an abandonment defense, the line between preparation and attempt may fall so early as to seem unfair to the defendant who voluntarily abandons his plan, yet fall too late to meet proper law enforcement objectives with respect to the defendant who apparently would have carried his plan through to completion.

One way to avoid this dilemma is to recognize that voluntary abandonment is a complete defense to a charge of criminal attempt. Although the common law decisions appear unsettled or in conflict with respect to the status of such a defense (Rotenberg, pp. 596–597), many statutory codifications have adopted it. For example, a New York statute (N.Y. Penal Law (McKinney) § 40.10(3)(1998)) provides a defense to an attempt charge when "under circumstances manifesting a voluntary and complete renunciation of his criminal purpose, the defendant avoided commission of the crime attempted" (cf. Model Penal Code, 1962, § 5.01(4)).

In jurisdictions that recognize an abandonment defense, it is necessary to determine when the abandonment is genuinely "voluntary." Given the rationale of the defense, it seems clear that abandonment should not be considered voluntary when prompted by realization that the police or the victim have detected the plan, or when the defendant is simply postponing the attempt until a more favorable opportunity presents itself. The Model Penal Code provides that "renunciation of criminal purpose is not voluntary if it is motivated in whole or in part, by circumstances, not present or apparent at the inception of the actor's course of conduct, which increase the probability of detection or apprehension or which make more difficult the accomplishment of the criminal purpose" (1962, § 5.01(4)).

Impossibility. Courts and commentators have struggled for generations over the question whether an accused should be punishable for attempt when, for reasons unknown to the defendant, the intended offense could not possibly be committed successfully under the circumstances. The problem arises in a great variety of settings. The accused, for example, may attempt to kill with a pistol that is unloaded of defective, or he may shoot at an inanimate decoy rather than at the intended victim. A would-be pickpocket may reach into an empty pocket, or a drug dealer may purchase talcum powder, believing it to be narcotics.

Some courts have sought to resolve such cases by distinguishing between "legal" and "factual" impossibility. Factual impossibility is said to arise when some extraneous circumstances unknown to the defendant prevents consummation of the crime, and in this situation the attempt is punishable. Legal impossibility, on the other hand, arises when the intended acts, even if completed, would not amount to a crime, and it is said that in this situation the attempt should not be punishable.

In application, these concepts of legal and factual impossibility have proved elusive and unmanageable. In one case involving a charge of attempt to smuggle letters out of prison without the knowledge of the warden, the plot was discovered by the warden, although the accused remained ignorant of this fact. The court treated the case as one of legal impossibility and reversed the conviction for attempt (United States v. Berrigan, 482 F.2d 171 (3d Cir. 1973)). It is apparent, however, that the situation could as readily be characterized as one of factual impossibility, and the same is true of attempts to pick an empty pocket, to shoot at a dead body believed to be alive, and so on. Some commentators have sought to clarify the categories by introducing further distinctions between "intrinsic" and "extrinsic" factual impossibility (Comment, pp. 160– 162). One court has suggested a still more sophisticated taxonomy involving six ostensibly distinct categories (Regina v. Donnelly, (1970) N.Z.L.R. 980, 990 (C.A.)).

All of these efforts at classification ultimately founder, however, because generally speaking the reasons for punishing unsuccessful attempts apply as much to one category as to any of the others. When the defendant has fired at a decoy or used an unloaded weapon, the circumstances may, of course, raise a question about whether he actually intended to kill, but the question of intent must be faced and resolved with care in every type of "possible" or "impossible" attempt. In fact, the use of undercover agents or cleverly disguised decoys may provide particularly reliable confirmation of intent, even though such tactics would arguably raise a problem of "legal impossibility" under some of the traditional taxonomies. So long as it can be proved that the accused acted with intent to commit the offense and that his conduct would constitute the crime if the circumstances had been as he believed them to be, the defendant is just as culpable and in general just as dangerous as the defendant who successfully consummates the offense. Nearly all of the modern statutory codifications have taken this view, specifying that neither factual nor legal impossibility is a defense "if such crime could have been committed had the attendant circumstances been as such person believed them to be" (N.Y. Penal Law (McKinney) § 110.10 (1998)).

There remains one type of "legal impossibility" that fails to satisfy the proviso just quoted. Suppose that the accused has attempted to smuggle expensive lace past a customs officer but that (unknown to the accused) this item has recently been removed from the lists of goods subject to duty. Here, even if the accused had accomplished everything he set out to do, his acts will not violate any provision of law. It is true that the accused thought he would be committing a crime, but since the goal he seeks to achieve is not in fact prohibited, the purposes of attempt law do not call for punishment (Kadish and Paulsen, pp. 362–366). In this type of situation, sometimes called a case of "genuine" legal impossibility, the attempt would not be punishable even under revised statutory provisions that otherwise reject both factual and legal impossibility as defenses.

Grading. Statutory provisions specifying the penalty applicable to a criminal attempt vary widely among American jurisdictions. A specific punishment may be provided for all attempts, or different penalty ranges may be specified according to the seriousness of the crime attempted. Under some statutes, for example, the maximum penalty is one-half that provided for the completed crime. Although a few states provide for the same maximum penalty for attempt and for the corresponding completed crime, this approach still appears to be the minority view; despite other variations in detail, in most jurisdictions an attempt will be punished much less severely than the completed crime (Model Penal Code, 1960, appendixes A and B to § 5.05).

What is the justification of this prevalent grading pattern? Relative leniency seems appropriate in the case of the defendants who have crossed beyond the domain of "mere preparation" but who nevertheless have yet to carry out every step that appears necessary to consummate the crime. When the attempt is incomplete in this sense, the intent and the dangerousness demonstrated are inevitably more ambiguous than when the actor has taken the decisive final step. Moreover, the lower penalty preserves some incentive for the actor to avert the threatened harm, even when he may be unable to meet the requirements for a complete defense of voluntary abandonment.

In contrast, the prevalent pattern of leniency for attempts appears difficult to justify when the defendant has carried out every step that appears necessary for successful completion of the offense, as, for example, when a defendant shoots at someone intending to kill, but the victim survives the wounds inflicted. In such a case the difference between a successful consummation of the crime and an unsuccessful attempt may result from fortuitous factors wholly beyond the control of the actor, and the sharp difference in applicable penalties appears anomalous.

It is sometimes suggested that the successful actor may be more dangerous or more culpable than the one whose attempt fails. Neither of these arguments can be considered valid over the general range of attempt situations (Schulhofer, pp. 1514–1517, 1588–1599). The Model Penal Code accepts that premise, and concludes that generally the maximum penalty for attempt should equal that for the completed crime. The Code provides, however, that in the case of the most serious felonies the penalty for attempt should be less severe than for the completed offense (1962, § 5.05(1)). The rationale for this limited exception to the general approach of equal treatment is that in this situation the use of severe sanctions can be minimized without impairing the deterrent efficacy of the law (Model Penal Code, 1960, commentary on § 5.05). The Model Penal Code rationale turns out to depend on a number of complex and problematic assumptions. Although the Code's goal of limiting the use of the most severe sanctions appears attractive, it proves difficult to show with any degree of generality that the Code's approach in fact has this effect; leniency for unsuccessful attempts may instead work to perpetuate unnecessarily severe and vindictively harsh sentences in the case of completed crimes (Schulhofer, pp. 1562–1585).

Intuitively, the most plausible explanation for more lenient treatment of attempts is that the community's resentment and demand for punishment are not aroused to the same degree when serious harm has been averted. This explanation, however, raises further questions. Can severe punishment (in the case of completed crime, for example) be justified simply by reference to the fact that society "demands" or at least desires this? To what extent should the structure of penalties serve to express intuitive societal judgments that cannot be rationalized in terms of such instrumental goals as deterrence, isolation, rehabilitation, and even retribution—that is, condemnation reflecting the moral culpability of the act? Conversely, to what extent should the criminal justice system see its mission as one not of expressing the intuitive social demand for punishment, but rather as one of restraining that demand and of protecting from punishment the offender who, rationally speaking, deserves a less severe penalty? Answers to these questions must be sought beyond the confines of attempt doctrine, for they reflect wider problems of democratic theory and normative political philosophy.