I. The Functions of a Criminal Code
II. An Evaluation System

A. Does the Code Contain a Comprehensive Statement of the Law's Commands?

1. Abolition of Uncodified or Common Law Offenses
2. Specification of Affirmative Duties and Justification Rules
3. Definition of Terms

B. Does the Code Effectively Communicate the Law's Commands to the Public?

1. Drafting Style
2. Organization

C. Does the Code Provide a Comprehensive and Accessible Statement of its Rules of Adjudication?

1. The Need for Comprehensiveness
2. Complete Specification of Necessary Rules
3. Sufficiency of Detail and Ease of Use

D. Does the Code Accurately Assess Who Does and Who Does Not Deserve Criminal Liability?

1. Appropriate Criminalization Decisions
2. Appropriate Liability Rules
3. A Comprehensive System of Defenses

E. For Those Offenders Held Criminally Liable, Does the Code Accurately Assess the Proper Grade of Liability and Punishment?

1. Consistency and "Grading" of Offenses
2. Recognizing Appropriate Distinctions

A. The Scoring System
B. Implementing the System

A. Comprehensiveness in Stating Rules of Conduct

1. Abolition of Uncodified and Common Law Offenses
2. Specification of Affirmative Duties and Justification Rules
3. Definition of Terms

B. Effectiveness in Communicating the Rules of Conduct

1. Drafting Style: Generally
2. Drafting Style: Simplicity of Justification Rules
3. Organization: Generally
4. Organization: Overlapping Offenses

C. Comprehensiveness and Accessibility of the Principles of Adjudication

1. Specification of Necessary Rules: Culpability requirements
2. Specification of Necessary Rules: Excuses and Nonexculpatory Defenses
3. Specification of Necessary Rules: Other Adjudication Issues
4. Sufficiency of Detail and Ease of Use

D. Accuracy in Imposing Liability

1. Appropriate Criminalization Decisions
2. Appropriate Liability Rules
3. A Comprehensive System of Defenses

E. Accuracy in Grading Liability and Punishment

1. Consistency and "Grading" of Offenses
2. Recognizing Appropriate Distinctions

F. Overall Ratings

V. Conclusion

Appendix A: Evaluation Form

Appendix B: Re-Scoring Evaluation Form

Appendix C

 

Each American jurisdiction has a criminal code. Most jurisdictions have substantially restructured and improved their codes since 1962, when the American Law Institute first promulgated its Model Penal Code. Such reform efforts are worthwhile, especially in criminal law, because many advantages flow from the thoughtful codification of criminal law rules. By compiling all criminal rules in a single comprehensive source, codification makes access to these rules easier, increasing the chance that citizens will know what the criminal law commands. A codified rule has the advantage of increased precision, which is likely to increase the uniformity of its application. Uncodified rules -- or, even worse, unenacted rules, such as common law offenses -- often suffer from vagueness and ambiguity, which increase the potential for bias, abuse, and arbitrariness in their application. The articulation of criminal law rules in an integrated code also makes it easier to detect internal inconsistencies and irrationalities among the rules, because clear and unified legislative expression makes the specific demands and effects of each rule, and the relationship among the rules, apparent in a way that expression through judicial opinions or uncoordinated legislation cannot. Finally, requiring statutory enactment of criminal rules assures that the criminalization authority remains the province of the legislature, as is desirable in a democratic society. The absence of a statutory provision to control a significant component of the criminal law amounts to a de facto delegation of criminalization power to the courts.

Crime and criminal justice are among the few perennially "hot" political issues to which legislatures are sure to pay frequent and close attention. But the virtues of codification are not always, or even usually, central to legislatures when they address these issues. No politician runs for election on a platform to "increase internal consistency within the criminal law." Thus, while criminal law attracts much legislative attention, criminal codes attract little. As a result, the advantages of codification commonly are realized only imperfectly in American codes.

Indeed, many American criminal codes are not true codes at all in the modern sense of cohesive, well-structured, and self-contained statutory schemes, but mere collections of statutory provisions similar to the generalized legislative "codes" of the last century and before. In many instances, even states that adopted modern, systematic criminal codes -- typically during the recodification wave of the 1960s and 1970s -- have since altered their codes through ad hoc amendment so as to make them dramatically less systematic and internally consistent. Often the deterioration of a code results simply from ignorance of the structure and operation that the code's original drafters intended. In other cases, the deterioration is the product of politics, where interest groups of one sort or another arrange special provisions of one kind or another.

Politicians bear few costs from enacting or developing poorly organized and drafted criminal codes, because the social costs of having such a code, even if substantial, are hidden and diffuse: uncertainty and confusion in prosecutions, and the lack of predictability and uniformity in adjudicative outcomes. The voting public would see these problems only if it understood the inner workings of the criminal justice system in a way that typically only judges, lawyers, and other frequent participants do. And these professional participants often are not sufficiently dissatisfied with their current situation to demand reform unless it is apparent to them that a better system is available. Moreover, these participants often have an investment in keeping the system as it is because, while it has its problems, at least its problems are known. Sometimes simple inertia keeps a bad code in place. Criminal code reformers often hear the complaints of judges, lawyers, and police officers that they don't want to have to learn new rules, or even new code section numbers. With little apparent cost to having a deficient criminal code and clear costs to undertaking a reform of it, legislatures are likely to pass up the virtues of recodification to keep the certainty and familiarity of the status quo.

Our hope in undertaking the project that resulted in this Article was to highlight the serious deficiencies (and strengths) of American criminal codes, and thereby to provide an incentive for reform. The Article develops a conceptual framework with which to evaluate the effectiveness of a criminal code, translates that framework into a quantitative scoring system, and uses the scoring system to rank the performance of all fifty-two American criminal codes.

Part I of the Article describes in general terms the functions criminal codes are meant to perform. Reasoning from these functions, Part II advances five general criteria for evaluating the effectiveness of a criminal code. Each of these five general criteria are dissected into a list of specific factors to guide assessment of a code on each of the five criteria. In Part III, we describe the scoring system we developed to apply the Part II criteria to the fifty-two criminal codes in the United States. Part IV reports the results of our evaluations. We provide a ranked list of the codes for each of the five general criteria and for overall performance and offer illustrations of the best and worst exemplars of the factors we used to calculate our scores. In its conclusion, the Article discusses the rankings and their implications.

I. The Functions of a Criminal Code

The criminal law has two primary functions. First, it has a rule articulation function: It must define and announce the conduct that is prohibited (or required) by the criminal law. Such "rules of conduct," as they have been called, provide ex ante direction to members of the community as to the conduct that must be avoided (or that must be performed) upon pain of criminal sanction.

It seems only reasonable that society tell its members in an understandable form what the criminal law expects of them. Indeed, our condemnation and punishment of criminals, as distinguished from civil violators, rests upon the assumption that a criminal violation requires some consciousness of wrongdoing, or at least a gross deviation from a clearly defined standard of lawful conduct. How can this assumption be sustained if the commands of the criminal law are unclear? How can we condemn and punish violations of the rules of lawful conduct if the general public does not, and cannot reasonably be expected to, know those rules? One also may wonder how effective the criminal law can be in deterring criminal conduct if the law's prohibitions are unclear. The criminal law thus has a great interest in effectively communicating its rules of conduct.

When a violation of the rules of conduct occurs, the criminal law takes on a different role, an adjudication function. The adjudication function has two components: the code must decide whether the violation merits criminal liability and, if so, how much. Thus the first adjudication issue, setting the minimum conditions for liability, assesses ex post whether the violation is sufficiently blameworthy "to warrant the condemnation of conviction." Where it has established that criminal liability is to be imposed, criminal law must address the residual adjudication issue of grading, assessing the general range of punishment that ought to be imposed. The degree of punishment is usually meant to reflect the moral blameworthiness of the offender. While the first step in the adjudication process (the liability function) involves a binary yes-or-no decision as to whether the minimum conditions for liability are satisfied, the latter step (the grading function) requires judgments of degree. It must consider such factors as the relative harmfulness of the violation and the level of culpability of the actor.

In performing these different functions, a criminal code addresses different audiences. In serving the first function, announcing the rules of conduct, the code addresses the members of the public. In performing the second and third functions, adjudication of a violation -- determining whether to impose liability and, if so, at what grade of liability and punishment -- the code addresses lawyers, judges, jurors, and others who participate in the adjudication process.

Because of these different audiences, the different functions often call for different drafting styles. To communicate the rules of conduct effectively to the public, the code must be easy to read and understand. It must give a clear statement, in objective terms if possible, of the conduct that the law prohibits and under what conditions it is prohibited. Readability, accessibility, simplicity, and clarity are the useful characteristics in this function.

The adjudicators, on the other hand, can tolerate greater complexity. Moreover, although clarity and simplicity are always virtues, the subtle and sophisticated judgments required of adjudicators necessarily limit the extent to which adjudication rules can be simple yet useful. While the public can be told rather easily and clearly that "you may not cause bodily injury or death to another person," when a prohibited injurious act does occur, the adjudicators need numerous and intricate rules to determine whether the injurer ought to escape liability because he or she had no culpability, was insane, mistakenly but reasonably believed that the force used was necessary for self-defense, or for any number of other reasons. If liability is appropriate, the grading rules must determine the degree of liability that is appropriate, taking account of the level of the actor's culpability, the extent of the injury, and a variety of other mitigating and aggravating circumstances. To properly perform these functions, many, if not most, of the liability and grading rules must use complex and sometimes subjective criteria.

II. An Evaluation System

Given the advantages of codifying criminal law (increasing fair notice of the law's commands, increasing uniformity in application, decreasing the potential for abuse of discretion, reserving the criminalization authority to the legislature) and given the special demands on a criminal code in performing its various functions (articulating rules of conduct and principles for adjudication of liability and grading a violation), how can one construct a system to evaluate the effectiveness of a criminal code?

It is important at the outset to be clear about what such an evaluation should not do. An assessment of a criminal code, as opposed to the criminal laws contained therein, should address itself to the integrity of the code as a scheme for articulating legal rules, not the wisdom of the rules the code articulates. In developing our evaluation scheme, therefore, we attempted to minimize the effect of our own value judgments about what should and should not be criminalized. We think this a matter for individual legislatures. Our focus is primarily in judging how well the legislature's criminalization decisions have been thought out, formulated, and communicated. To avoid substituting our own value judgments, we were critical of the substance of codes only in those situations in which a code diverged from widely shared intuitions of justice that were dramatic and generally uncontested -- that is, instances where the disparity seems inadvertent (the drafters themselves probably did not intend the rule's result) or where the injustice of the rule is acknowledged but justified on some other ground (for example, the drafters claim that a less unjust rule cannot be drafted or would entail undesirable practical problems that other jurisdictions have shown can be overcome).

We developed two criteria by which to evaluate a code's effectiveness in announcing the rules of conduct. First, the code must be comprehensive in describing the rules of conduct. Second, it must communicate those rules effectively to the general public. Fuller explanations of these two criteria, and of the factors we used to judge a code's satisfaction of these criteria, appear in Parts II.A and II.B of this Article, respectively.

We also formulated three criteria by which to judge a code's adjudication rules. First, the rules for both liability and grading must be comprehensive and accessible. Second, the liability rules should capture the community's sense of justice. Phrased negatively, the question here is: would application of the code's adjudication rules regularly result in criminal liability for persons that the community would agree do not deserve it? Third, the code must establish accurate and consistent rules for determining the grade of a violation. In other words, application of the code's grading provisions should not regularly impose an amount of punishment that the community would perceive as unjust (either too much or too little punishment). These three criteria, and the factors relevant to their evaluation, are elaborated and defended more thoroughly in Parts II.C, II.D, and II.E.

A. Does the Code Contain a Comprehensive Statement of the Law's Commands?

A code must tell persons what conduct it prohibits, what conduct it requires, and what conduct normally prohibited is permitted -- justified -- under special conditions. Obviously, a code cannot inform citizens of the law's commands if it does not include all of them -- in other words, if it fails to codify some offenses and justification defenses. Codifying an offense or justification defense but leaving important terms undefined can have a similar effect. Failure to include or fully explain the nature of a criminal offense, an affirmative duty, or a justification rule essentially leaves the criminalization decision to the courts and makes it difficult, if not impossible, for a layperson to know the rules of conduct. An elaboration of the factors we used to judge the comprehensiveness of criminal codes follows; a summary list of these factors appears under Question 1 of our Evaluation Form.

1. Abolition of Uncodified or Common Law Offenses

A code that is not self-contained, but admits of the possibility of criminal punishment for uncodified offenses, cannot provide proper notice or assure consistent enforcement. Uncodified offenses may be generated in two ways: the legislature may set out additional crimes that appear in statutes other than the criminal code, or the courts may define "common-law crimes." A code that fails to foreclose both of these possibilities by way of an explicit prohibition will have at least three shortcomings.

First, and most obviously, the likelihood of notice to the populace at large diminishes as the dispersion of criminal provisions in the state's statutory scheme increases. It is simply much easier for the layperson to educate herself about the state's criminal law if that law can be found in one place. In fact, a state that codifies some but not all crimes may present a particular problem: To the extent that a seemingly unified "criminal code" will lead citizens to believe that all crimes are included therein, the existence of additional criminal provisions outside the code will frustrate citizens' expectations.

A second, and subtler, "notice" problem will affect the legislature itself. If crimes are spread throughout the state statutory code, the legislature will be less likely to view the criminal law as a consistent, unified scheme. A new offense may be placed outside the code, making it less likely that the legislature will consider how that offense fits within the existing matrix of criminal offenses. Additionally, the criminal code itself may be amended without consideration of the amendment's impact on offenses outside the code.

Third, the presence of criminal offenses outside the code will likely generate problems of statutory construction. For example, it may not be clear whether the legislature expected the criminal code's "default" culpability provision to apply to uncodified offenses. In short, the possibility of criminal offenses appearing outside the criminal code undermines the entire project of setting aside a separate criminal code within the overall state code in the first place.

Failure to eliminate the possibility of common-law offenses is a still greater fault for a criminal code, for they create a more significant notice problem. The average citizen simply cannot be expected to peruse a state's case law to ferret out non-statutory offenses, and certainly not earlier English case law. In addition, judicial authority to "define" offenses raises serious democracy concerns. It is essential that the criminal law be rooted in legitimate moral consensus; otherwise the law will be unsatisfactory both on normative grounds and in practical or utilitarian terms. As a normative matter, the law will not track the community's true sentiments with respect to what conduct is considered blameworthy. As a practical matter, a code that does not accurately parallel community norms will fail to deter some objectionable conduct; it will over deter some unobjectionable conduct; and (because its moral sanction will not be taken seriously by the regulated population) it will under deter even the "bad" conduct it proscribes. Though the legislative determinations may at times be a poor proxy for genuine community consensus, they are doubtless superior in this respect to judicial determinations.

2. Specification of Affirmative Duties and Justification Rules

A criminal code must communicate not just what persons may not do, but also the conduct that a person must perform on pain of criminal liability -- that is, conduct mandated by a legal duty -- and the conduct that a person may perform under certain conditions although it otherwise would be prohibited -- that is, conduct that would be criminal but for a justification defense. A code that fails to make clear the duties that can give rise to criminal liability, or fails to define fully the conditions under which otherwise criminal conduct is permitted, does not adequately advise citizens of the rules of conduct the criminal law sets for them.

The need for a comprehensive statement of the rules of conduct applies no less to the criminalization of a failure to act than to affirmative conduct offenses. Yet even the Model Penal Code -- which forbids the creation of common-law offenses -- provides that courts, in addition to legislatures, may define such affirmative duties upon which criminal liability may be imposed. There seems no principled basis for distinguishing omissions from acts in this way. If anything, it is more important that the legislature specify with precision the situations in which failure to act may lead to criminal liability. This is particularly true given our law's general reluctance to hold people legally accountable for their omissions. The default expectation in a system such as ours is that doing nothing at all will not raise the possibility of criminal punishment. Of course, there may well be situations where the community's moral consensus holds that one cannot stand idly by, but is obliged to do something or risk being branded a criminal. It should be incumbent on the legislature, however, to notify the public as to these obligations by way of an explicit codified provision. Accordingly, it is at least as improper for a code to allow courts to define common-law "affirmative duties" of criminal moment as to allow them to define common-law offenses.

The same notice concerns apply to the codification of justification defenses. Absent a clear delineation of the situations in which otherwise prohibited conduct will be exonerated, the criminal code will over deter and will discourage people from engaging in conduct that the law is happy to have performed, and indeed may well wish to encourage. Notice is thus as significant with respect to justification defenses as it is to the definition of the contours of an offense in the first place. Indeed, delineation of justification rules is best seen merely as a clarification of the true definition and scope of the conduct rules; no code of conduct can be deemed comprehensive if it omits this necessary refinement.

3. Definition of Terms

To be truly comprehensive, a code must be not only complete, but precise; code provisions "covering" proscribed behavior are useless if they do not clearly define the nature of the proscribed behavior. For example, a code with a provision that simply reads, "theft is prohibited," is surely complete in its prohibition of theft -- whatever that may be -- but is so imprecise that it provides little useful information. In such a situation, it becomes the courts' responsibility to ascertain the parameters of the code's meaning. One cannot understand such a provision without reference to the case law that interprets it, and the objective of comprehensiveness is not satisfied. Therefore a truly comprehensive criminal code must sufficiently define all relevant terms so that reference to outside sources is unnecessary.

B. Does the Code Effectively Communicate the Law's Commands to the Public?

Even if a criminal code were perfectly comprehensive in setting out its rules of conduct -- defining prohibited conduct, required conduct, and the conditions under which otherwise prohibited conduct is permitted -- it would fail in its function if it did not present those rules of conduct in a way that lay persons could understand. Effective communication demands clarity at two levels: (1) within each rule, and (2) in the organization of the rules into a code. These factors are described more thoroughly below and are summarized on our Evaluation Form under Question 2.

1. Drafting Style

The criteria relevant to a determination of the effectiveness of the communication of a rule are obvious enough. The style of drafting is important. Offenses ought to be drafted in a way that makes the provisions understandable to lay persons. Thus, the rules of conduct should use common and plain words where possible, and provide straightforward definitions for other words. The sentences comprising each provision should be as short and clear as is feasible. Informative section titles also can be useful. Each provision should bear a title or heading that accurately summarizes the general nature of the conduct addressed therein.

There is one specific category of provision for which simple, effective communication of the rules of conduct is particularly important: justification defenses. Criminal offenses relate most commonly to behavior that one may plan to do or not do in advance. At least in theory, then, one has time to review even a reasonably complex criminal offense provision beforehand to determine whether one's planned conduct is a crime. Justifications, on the other hand, relate not to actions but reactions. They prescribe rules that govern what one may or may not do in situations where one has only a moment to decide what she will do (and, frequently, where it is imperative that one do something). For this reason, it is crucial that justification rules be clear and simple enough that one might implement them accurately at a moment's notice, without time for reflection or deliberation.

2. Organization

In addition to expressing each of its rules clearly, it is critical that the code organize the rules effectively and sensibly. First, the code should group similar offenses into categories to facilitate easy location. Listing offenses alphabetically would seem a straightforward and simple method, but is not effective because offense nomenclature is not always known in advance (for example, the same crime may be called "murder" or "homicide," "theft" or "larceny"). In other words, alphabetical organization works only if one assumes the lay reader already knows the official legal name of the offense. Further, the distinctions between somewhat similar offenses (for example, theft versus robbery versus burglary) often become clear only when the offenses are viewed together, especially given the possible problem of a code's containing overlapping offenses, of which we shall speak shortly.

Within offense categories, it is best to list offenses in order of seriousness, either increasing or decreasing. This method makes the most serious offenses easy to find, and, more importantly, serves the important expressive function of conveying to the layperson which offenses society deems most serious. In many situations, citizens must consider not only the existence, but the relative magnitude, of the moral sanction associated with particular acts. In fact, some rules of criminal liability -- such as the "lesser evils" defense -- rely directly on citizens' ability to discern and compare the relative moral stigma attaching to different activities.

One common organizational fault of criminal codes is that they thoroughly mix the rules of conduct with the principles of adjudication, so that a relatively simple conduct prohibition becomes lost in a mountain of complex provisions for adjudicating violations of the prohibition. The most egregious form of this flaw is the failure to distinguish between a code's General Part, which sets out the principles of general application, and its Special Part, which contains the specific offenses.

The presence of offense definitions that "overlap" to penalize the same conduct in two or more offenses creates problems beyond mere sloppiness. Overlapping offenses reduce the effectiveness of notice, and the ease of adjudicative application, because they confuse the reader. For example, how ought one read an offense provision that specifies as criminal an activity already apparently covered by another, more general provision? Given the general canon favoring a statutory reading that avoids superfluity, one would have to assume that the general provision did not in fact cover the more specific conduct. This in turn would suggest that the reach of the broader statute is in fact more limited than had been surmised, affecting interpretation of that offense in other areas as well. Unnecessary "add-on" offenses also reduce the moral and practical force of a criminal code. Simply put, the more provisions a code has, the less effective each provision is likely to be. The moral gravity of committing a crime weighs less heavily as the code criminalizes more conduct -- particularly since at the margin, each new crime presumably prohibits less blameworthy conduct than any pre-existing crime. Moreover, true notice becomes less plausible a citizen is expected to keep an increasingly large number of possible offenses in mind at once.

C. Does the Code Provide a Comprehensive and Accessible Statement of its Rules of Adjudication?

Just as a criminal code must be comprehensive in its articulation of the rules of conduct, comprehensiveness is equally important for the code's elaboration of rules of adjudication, although for somewhat different reasons, as set out below. A summary of the factors we used to evaluate this criterion appears under Question 3 of our Evaluation Form.

1. The Need for Comprehensiveness

Undefined adjudication rules vest unguided discretion in decisionmakers, which can breed disparity in application and create the potential for abuse. The courts may ultimately supplant the missing legislative rule with a judicial rule, but this effectively amounts to a legislative delegation of the substantive criminalization decision. Moreover, development of an authoritative judicial rule takes time to occur (assuming it ever occurs), and similar offenders having committed similar offenses may experience arbitrary variation in liability determinations during the interim. Of course, there are unavoidable limitations on the legislature's ability to prescribe principles covering all adjudication decisions in advance. Some situations are simply unpredictable. And in these cases, the system must rely upon courts to provide an adjudication rule for the case at hand. But a code ought to at least make an effort to state the rules that it can, especially those typically needed for the adjudication of unexceptional cases.

While both adjudication rules and conduct rules share a need for comprehensiveness, the two serve different functions, and adjudication rules are subject to a different set of problems than rules of conduct. Accordingly, a different drafting approach is needed for adjudication provisions. Greater detail often is desirable in adjudication rules because such increases uniformity in the adjudication of similar cases. Greater complexity is tolerable because decision makers will have legal training that the lay persons applying conduct rules will not. (Even juries get detailed jury instructions that give them the law governing the case.) Adjudicators also have time for thoughtful reflection in applying the provisions. Additionally, providing the average citizen with notice of the code's rules of adjudication is considerably less important than providing notice about prohibited conduct.

2. Complete Specification of Necessary Rules

The need for comprehensiveness becomes clearer when we consider specific elements of the code's rules of adjudication. First, the code must contain, and define, the culpability requirements that will determine whether improper conduct deserves criminal sanction. Without an exhaustive list of culpability requirements, the relationship between those requirements, and the manner in which the requirements will be applied to the rules of conduct, a criminal code has done only half its job: it has codified each actus reus without any explanation of the corresponding mens rea necessary for criminal punishment.

Additionally, no set of adjudication rules will have accomplished its purpose if it ignores certain practical elements that bear on the connection between conduct and liability. We shall refer to the rules governing these elements as general adjudicatory provisions. For example, the issue of causation, though evidentiary in part, is nevertheless critical to the determination that an actor's conduct amounts to a crime. The criminal code should articulate the substantive standard governing when an act is the legal cause of a result. Similarly, a code must include provisions explaining the effect, if any, of a "victim's" consent to an otherwise criminal act.

A code must also include a clearly defined exposition of the contours of each defense that will excuse improper conduct. Vesting the courts with unfettered discretion to determine what counts as "insanity" or "duress" -- or, worse, to determine whether these excuses will be available in the first place -- would undermine the goal of consistency and would provide a delegation of authority that ought to be reserved for the more democratic legislature. Similarly, a code must address certain mental states that, although they do not amount to excuses, nevertheless bear on the extent to which an actor should be subject to criminal sanction. For example, although an offender who voluntarily becomes intoxicated before committing a crime will not be totally excused for that reason, his altered mental state may well mitigate the proper extent of his criminal liability. Finally, the same holds true (indeed, probably even more so) for nonexculpatory defenses, that is, the rules that prevent imposition of criminal liability even where the offense definition is satisfied and no justification or excuse defense is available, such as "statutes of limitations."

Finally, a truly complete code will provide the courts with a clear general message as to how they should read the code's provisions. Ideally, the code's provisions will be so clear that the judiciary will have little trouble applying them in specific cases. Yet ambiguities, doubts, and unanticipated situations inevitably arise, and the courts will have to interpret the code's meaning in situations where its language gives incomplete guidance. The legislature should anticipate this eventuality and offer some general prescription regarding the manner in which courts should interpret the code. For example, even as it tries to avoid creation of overlapping offenses, a code should provide rules to guide the adjudication of conduct that might fall within several offense definitions at once.

3. Sufficiency of Detail and Ease of Use

Of course, as with the rules of conduct, the mere presence of a rule of adjudication does not help if it is not sufficiently thorough. Indeed, if anything, more thoroughness is required here -- or at least, more rigorous language is needed to achieve the goal of thoroughness. Whereas simplicity and economy are crucial for conduct rules, there is a certain virtue in detail as regards the elaboration of adjudication rules. The goal of uniformity in application requires that the code make clear which seemingly different offenders merit punishment in equal measure and which superficially similar offenders must be treated differently from one another. Similar cases must be treated equally, and different cases must be distinguished according to well-defined principles. Increased complexity does not pose a problem for the rules of adjudication -- they will be applied ex post by a neutral adjudicator who may give them her full attention, rather than needing to be known ex ante by all citizens all the time. Greater detail serves to confine the adjudicator's discretion and focus her attention on relevant considerations rather than allowing her to be swayed by unimportant concerns.

D. Does the Code Accurately Assess Who Does and Who Does Not Deserve Criminal Liability?

A central task of a criminal code is to impose punishment on those who deserve the condemnation of criminal liability and to protect from punishment and condemnation those who do not deserve it. Determination of a code's success in accomplishing this task demands the most subjective value judgements of any of our criteria, for it often verges on an outright assessment of the normative propriety of the legislature's substantive decisions about what to include in the criminal code. Still, it is impossible to evaluate the utility or effectiveness a code without undertaking some inquiry as to what it does and does not contain. Frequently this inquiry operates only at the surface of the code rather than delving into a sweeping examination of its substance. For example, one can ascertain whether the code establishes minimum culpability requirements for every offense without evaluating the specific culpability level chosen for each offense. Sometimes, however, substantive decisions impact so fundamentally on the efficacy of the code -- as a code, rather than a mere assemblage of rules that may or may not have merit standing alone -- that to evaluate the code's integrity is to evaluate those specific decisions.

This inquiry takes place on two levels. First, it is necessary to review the code's rules of conduct to ensure that they satisfy a minimum threshold of acceptability in deciding what conduct to criminalize. We discuss this element of the analysis in section II.D.1. Second, the evaluation must undertake a more searching analysis of the code's rules of adjudication. This analysis itself has two components: the code must establish appropriate minimum culpability requirements that determine when violations of the conduct rules shall be deemed criminal; and it must include all appropriate defenses to avoid imposing liability where an otherwise criminal act is justified or where an otherwise criminal actor is excused. We discuss these two components in sections II.D.2 and II.D.3, respectively. Question 4 of our Evaluation Form summarizes the specific factors used in evaluating this criterion.

1. Appropriate Criminalization Decisions

The first tier of the analysis of the accuracy of the code's liability assessments relates to the rules of conduct: does the code criminalize suitably "bad" behavior and only such behavior? Obviously, a code whose commands horribly fail to mesh with the moral consensus of the regulated community will lack credibility. A legislature can commit no more grievous fault when enacting a criminal code than fundamentally to "get it wrong" when deciding what behavior to criminalize. Of course, many criminalization decisions will reflect subjective value judgments as to which different communities might reasonably differ, so it would be mere conceit to evaluate each provision of a code based on one's own sense of its moral rectitude. One can, however, impose at least two basic minimum requirements in this regard: a code should not include provisions that outlaw patently trivial moral transgressions or that are incapable of enforcement.

The enactment of "trivial offenses" is itself far from trivial. Such provisions undercut the moral force of the criminal law and may have potential spillover effects on law enforcement or prosecution. Some people might cite prosecutorial discretion as a panacea for any legislative overreaching; but such discretion is as likely to exacerbate as to counteract the dangers of over-criminalization and, in any event, blind reliance on discretion at any level only opens the door to the type of selective, disparate treatment that adjudication rules should combat.

Two specific code provisions can combat the possibility of imposing liability for insufficiently serious acts. First, the code might (following the Model Penal Code) include a general defense against punishment for a de minimis criminal infraction. Second, the code might (rejecting the Model Penal Code) define a rigorous standard for imposition of liability for attempted offenses. Such a definition of attempt would prevent imposition of liability on acts that might never have posed a legitimate social danger or on actors who remained likely to change their minds before following through on their initial criminal intentions. By contrast, the Model Penal Code -- which imposes liability for any activity that constitutes a "substantial step" toward the commission of the completed offense -- has a relatively weak and nebulously defined attempt requirement. Public sentiment is strongly in favor of a more demanding attempt requirement. Use of the lesser standard reduces the authority of the criminal code.

2. Appropriate Liability Rules

The second tier of the analysis under our fourth criterion examines the adjudication provisions. This examination looks to whether the actor, as opposed to the act, is sufficiently "bad" to merit criminal sanction. The first element of this review asks whether the code's adjudication rules set appropriate culpability requirements. Accurate reflection of societal norms regarding the criminal sanction demands that only the blameworthy be punished for committing bad acts. To ensure that only blameworthy actors will be punished, the code must establish minimum culpability thresholds. The easiest and best way to ensure that proper minimum threshold has been set for each offense is to provide a culpability level that the adjudicator may automatically "read in" as an element of an offense that is otherwise silent as to its culpability requirement. Such a provision guarantees that an overzealous adjudicator will not impose a strict liability or negligence standard where such a standard would be inadequate. More particular problems arise when codes include offenses that, by definition, fall below any proper culpability threshold. For example, a generalized felony-murder rule improperly assumes a degree of malice in taking human life under a broad array of circumstances.

3. A Comprehensive System of Defenses

The second element of the analysis of adjudication rules asks whether the code contains all appropriate exculpatory defenses. As noted above, a code that fails to include a defense leaves the definition of that defense to the courts. In this situation, not only is the properly legislative function of defining the scope of the defense delegated to the judiciary, but the courts may very well reach the unsatisfactory conclusion that the code's silence is meant to declare that the defense is not available -- a particularly plausible statutory gloss for codes that define some defenses but not others.

It is imperative, then, that a code include all appropriate defenses and leave nothing to the whim of the judiciary. Many of these defenses are uncontroversial, including both justification defenses that make otherwise unacceptable acts acceptable (covering acts in self-defense or defense of others and acts undertaken in a law enforcement capacity or while serving in a similar position of "special responsibility") and excuse defenses that negate culpability (insanity, immaturity, and involuntariness, including actions following involuntary intoxication). Two other defenses are less common and more frequently contested, but are also appropriate and desirable: the lesser evils or necessity justification defense and the rare, but appropriate, excuse defense of a reasonable mistake of law (whether based on the law's unavailability or an actual misstatement of the law in an official document).

E. For Those Offenders Held Criminally Liable, Does the Code Accurately Assess the Proper Grade of Liability and Punishment?

The second part of the adjudication task is the determination of how much liability and punishment ought to be imposed on an offender. A criminal code's role in this process is to set the general rage of punishment, with the sentencing judge, often guided by sentencing guidelines, finishing the task by fixing the amount of punishment at a particular point within the range that the criminal code's grading judgment has identified. Sentencing judges also will make the decision as to the method, as opposed to amount, of punishment: prison, house arrest, supervised release, fine, community service, or some other form of punishment.

Evaluation of a state criminal code's accuracy in determining the appropriate grade of liability and punishment for a given offense employs two rather straightforward standards: (1) the consistency standard, which asks whether the code grades similar offenses similarly and with respect to their relative seriousness; and (2) the differentiation standard, which asks whether the code recognizes all of, and only, the proper distinctions between different "degrees" of criminal conduct. The consistency standard demands that offenses "equal" in gravity (though potentially vastly different in terms of the conduct they proscribe) be treated alike by the code; the differentiation standard demands that offenses differing in gravity (though substantially very similar in terms of the conduct they proscribe) be treated suitably differently. The factors guiding evaluation of a code's satisfaction of these two standards are summarized under Question 5 of our Evaluation Form and are discussed more fully in sections II.E.1 and II.E.2.

1. Consistency and "Grading" of Offenses

Determination of the extent to which a code satisfies the consistency standard requires an examination of the overall system for assessment of punishment levels under the code. A critical factor here is the presence, and the sophistication, of a general "grading structure" for offenses. For example, a code might define seven grades of felonies and three grades of misdemeanors, each of which are subject to specific minimum and maximum sentences or fines. Obviously, however, where codes do not grade offenses, but rather provide specific sentences for each offense, an examination of the relative appropriateness of the sentences themselves becomes necessary.

A grading scheme with only a few categories essentially delegates most of the grading task to the sentencing judge. Few very broad grades provide great judicial discretion, which undercuts uniformity in application and increases the potential for abuse of discretion. Sometimes sentencing guidelines will provide the structure and limits on discretion that a criminal code's grading scheme fails to provide. But shifting the grading task to sentencing guidelines also is problematic. Such a shift takes the jury, as well as the procedural safeguards of trial, out of the process of determining the facts upon which will be based the determination of the amount of punishment to be imposed.

At the most fundamental level, the overall complexity of a grading scheme serves only as a proxy for a deeper review of the appropriateness of each of the code's specific punishments. Even so, a system of grading categories can be said to have its own practical value, for it forces the legislature to consider the relative seriousness of an offense vis-a-vis other offenses when assigning a grade to that offense. Clearly, the fewer grading categories there are, the more inexact (and inaccurate) these legislative determinations, and the adjudicator's subsequent punishment measures, will become. The absence of any grading scheme whatever necessarily makes review of the code's consistency more difficult and enhances the likelihood that legislative determinations respecting suitable levels of punishment are being made on an ad hoc basis by offense.

Obviously, in addition to a review of the grading system of a state criminal code, examination of a code's consistency demands a review of the implementation of that system. The creation of a large number of grading categories is useful only insofar as it enables the legislature to tailor punishments to crimes, grouping offenses of similar seriousness together. Thus the consistency standard also inquires whether the offenses in each grade are actually roughly equivalent in seriousness or, where the code created no grading system, all of the offenses for which the same sentence is established are actually roughly equivalent in seriousness.

2. Recognizing Appropriate Distinctions

In evaluating codes' satisfaction of the differentiation standard, it becomes important to look not at the overall punishment scheme, but at the distinctions made within specific sets of offense categories. For each category, the code should recognize all appropriate aggravating and mitigating conditions when specifying different "degrees" of an offense within the category. The code must establish the proper weight of each relevant factor, and the relation of each factor to other factors. Obviously, the relevant conditions vary from category to category: first-degree and second-degree sexual assault are (or should be) governed by different distinctions than first-degree and second-degree larceny. One mitigating factor relevant to all offenses, though -- and frequently identified in a general provision of the code for that reason -- is the absence of any resulting harm, i.e., the distinction between a completed and an inchoate offense.

This, then, sketches the factors that we looked to in evaluating a criminal code. Specific illustrations of good and bad code characteristics appear in Part IV and will illuminate our more abstract discussion here. But before turning to those results, we describe the procedures by which we developed and applied our evaluation scheme.

To develop a scoring system, we first attempted to articulate the general goals a criminal code should seek to fulfill, as described in Part I. We then reduced those general goals to a more specific set of five criteria, described in Part II. Using these criteria, we drafted an evaluation form to use as a guide to scoring. This draft form was tested on many codes and refined, then retested and further refined, until we had crafted a final version of the form. Below we detail the scoring method reflected on the form and the process by which we used the form to arrive at scores for the fifty-two American criminal codes.

A. The Scoring System

The form asks five "questions" that track our five criteria. For each question, a criminal code is given a score between 0 and 4. A score of "0" would mean that the code qua code was entirely useless; that is, it did not satisfy any of the specific objectives of a criminal code. A state with such a code would be no better (with respect to that criterion, at least) than a state whose criminal law was a complete mishmash of uncollected statutes, or relied entirely upon common law rules. A score of "4," on the other hand, would indicate a code that could not reasonably be expected to be improved upon in its performance of the functions captured by that criterion.

To assist in scoring according to this 4-point scale, we developed two general guidelines. First, we devised general descriptions of the qualities of a code that would receive a score of 0, 1, 2, 3, or 4 for each question, and included these descriptions on the Evaluation Form. Deciding which description best matched the reviewer's sense of the code's overall performance on a given question would aid in the determination of a final score. Second, after developing the specific scoring rules discussed in the following paragraph, we collectively scored the Model Penal Code's performance on the 4-point scale for each question. This offered one clear benchmark for scoring other codes. Since the Model Penal Code scored well on most of the questions, this benchmark was far more useful for good codes than for bad ones. It often provided very useful guidance for that subset of the codes. Because many codes follow the Model Penal Code in many respects, any deviations could be highlighted and compared to the Model Code, and a code's score could be revised upward or downward accordingly.

In addition to these general standards, we developed several specific bright-line scoring rules and included them on the Evaluation Form. These rules most often took the form of restrictions on the maximum score a code could receive if it failed to satisfy certain significant, and objectively ascertainable, specifications. The rules are summarized in the margin. By way of justification for these rules, we refer the reader to our elaboration of the concerns supporting the use of the factors to which the rules apply. Additionally, we underscore the fact that these rules all relate to factors sharing three characteristics: (1) each is significant, addressing a fundamental (if not the fundamental) concern of the question to which it applies; (2) each is pervasive, such that the impact of a flaw will almost surely spill over, directly or indirectly, onto the code's ability to satisfy other factors under the same criterion; and (3) each is objective, thus offering clear guidance to the scorer and making scoring more consistent. Because the rules relate to factors with these three qualities, it is almost certain that any code violating a given rule would receive almost exactly the same ultimate score even if the rule did not exist, that is, even if our scoring only used broad standards instead of incorporating these bright-line rules.

B. Implementing the System

Using the evaluation form reproduced in Appendix A, each of the fifty-two American codes was scored by two scorers working independent of one another. Discrepancies in the scoring on any of the five general questions were discussed and resolved at a series of group meetings of all members of the project. Once all discrepancies of more than 0.5 were resolved, sometimes requiring additional research into the codes, all fifty-two codes were ranked for each question.

This first tentative ranking of codes for each question was then used as the starting point for rescoring the codes. During this rescoring phase, a single scorer rated all fifty-two codes but only for a single question, as a means of more carefully testing whether each code's score for that question placed it in the proper rank-ordered position relative to other codes.

After this second scoring round, a number of "quality control" measures were executed as a final means of ensuring accuracy and consistency in our scoring. The scorer for each question transcribed, re-verified, and expanded his or her notes for each factor of the question he or she had scored during the second stage. The notations for each of our five questions for each of the fifty-two codes generated a database of 260 "paragraphs" (though not written in full sentences or, indeed, even in proper English at times) which would either justify a scorer's decisions or make his or her analytical missteps or omissions obvious.

To ensure that the more nebulous, less quantifiable factors were given their due and that codes' performances with respect to those factors were evaluated and compared consistently, we also developed a set of numbered "baskets" for each factor and required scorers, when finalizing their paragraphs, to place each code into one of these baskets for each factor. A code's basket placements would provide a convenient, if rough, shorthand method for broad-based comparisons between a code and other codes. Additional notes for each factor would enable recognition of more subtle distinctions in order to "fine-tune" the scores for codes ranked near one another. This more elaborate evaluation form, giving a set of alternative categorization "baskets" for each factor of each question, is reproduced in Appendix B.

To complete the quality-control process, each scorer's paragraphs were handed over to a reviewer who used the basket placements and additional notes to assess the accuracy of the scorer's code rankings. Where the notes seemed incomplete, the reviewer could ask the scorer to provide additional information to supplement the existing paragraph. Where the rankings seemed inappropriate, the reviewer asked the scorer for a justification of the seemingly anomalous scoring, and the reviewer and the scorer then worked together to arrive at a mutually satisfactory -- and, at long last, final -- score for every code on every question.

These final scores appear in the ranking tables presented in Part IV. The documentation paragraph for each of the five questions for each of the fifty-two jurisdictions are reported in Appendix C.

The performance of the codes varied greatly, both in terms of the codes' scoring relative to one another for each criterion and in terms of any particular code's performance across the various criteria. Close study reveals a remarkable variety of legislative approaches and decisions in addressing the task of fashioning a criminal code. Nonetheless, some general trends emerged both for each question and with respect to scoring among the different factors of a single question. Frequently, the codes in the top or bottom scoring range for each question had a great deal in common. Sections A through E of this Part analyze general scoring trends on each question and provide concrete examples of the successes and shortcomings of American criminal codes with respect to each of our scoring criteria. Section F discusses the codes' total scores for the five questions combined, and analyzes overall score trends.

A. Comprehensiveness in Stating Rules of Conduct

The codes' scores on question 1 are as follows:

1. NJ 3.75	27. CT 2.4
2. KY 3.7	28. NY 2.3
3. DE 3.65	29. OR 1.7
(MPC 3.65)	30. USC 1.55
4. NH 3.65	31. SD 1.3
5. AK 3.55	32. NV 1.25
6. AL 3.55	33. OK 1.25
7. AZ 3.55	34. OH 1.15
8. CO 3.55	35. WY 1.15
9. HI 3.55	36. CA 1.1
10. MN 3.55	37. NE 1
11. MO 3.5	38. WA 0.7
12. PA 3.45	39. ID 0.6
13. AR 3.4	40. NM 0.6
14. ND 3.4	41. FL 0.55
15. IN 3.35	42. MA 0.55
16. MT 3.35	43. MD 0.55
17. IA 3.3	44. MS 0.55
18. LA 3.3	45. SC 0.55
19. WI 3.3	46. VT 0.55
20. TX 3.25	47. VA 0.5
21. UT 3.2	48. MI 0.45
22. GA 3.1	49. DC 0.4
23. IL 3	50. NC 0.4
24. KS 3	51. RI 0.4
25. ME 3	52. WV 0.3
26. TN 2.5

 

Relatively speaking, the codes' performance on this question was solid but hardly spectacular. Although the high scores were among the highest for any question -- as evidenced by the average of the codes in the top quartile (3.569) and in the top half (3.381) -- there were also numerous states with very low scores. This is somewhat alarming, as a criminal code that scores especially poorly on this question is in many ways hardly a "code" at all, but is rather akin to a patchwork full of holes. Below we consider states' performances on each of our scoring factors.

1. Abolition of Uncodified and Common Law Offenses

Four of the five lowest-ranked codes under this question are silent as to whether common law offenses can be prosecuted. The fifth, Rhode Island, explicitly authorizes the prosecution of common law offenses:

Every act and omission which is an offense at common law, and for which no punishment is prescribed by the general laws, may be prosecuted and punished as an offense at common law. Every person who shall be convicted of any such offense which is a misdemeanor at common law shall be imprisoned for a term not exceeding one year or be fined not exceeding five hundred dollars ($500). Every person who shall be convicted of any such offense which is a felony at common law shall be imprisoned for a term not exceeding five (5) years or be fined not exceeding five thousand dollars ($5,000).

The possibility of significant punishment (here, up to five years' incarceration) for an uncodified (indeed, unenacted) crime severely undercuts the usefulness of having a criminal code at all.

On the other hand, the highest-ranked codes explicitly abolish common law offenses. The five top codes all have provisions similar to section 1.05(1) of the Model Penal Code, which bars the prosecution of offenses not defined by the criminal code or another statute. Thus, these criminal codes limit criminal liability to those offenses whose elements are defined and prescribed under the code, fulfilling the notice function.

2. Specification of Affirmative Duties and Justification Rules

The five lowest-ranked codes on question one -- along with the vast majority of the other criminal codes -- provide no guidance whatever regarding the existence or nonexistence of affirmative duties. In contrast, four of the five codes receiving the highest scores on question one -- New Hampshire, New Jersey, Delaware, and North Dakota -- deal explicitly with the subject of affirmative duties. Delaware and New Hampshire require that affirmative duties be defined by statute, though not necessarily by the criminal code. For example, Delaware's definition of "crime" provides that "'[c]rime' or 'offense' means an act or omission forbidden by a statute of this state." North Dakota and New Jersey have specific provisions governing the matter but are not useful; they essentially track the Model Penal Code, which allows for the possibility of criminal liability for affirmative duties defined by common law.

The lowest-ranked states under question one have no general justification sections. In fact, North Carolina is the only state of the lowest five to include any justification defense at all: a specific self-defense provision justifying "use of deadly force against an intruder." This represents a profound failing in the codes, for even a bad provision is better than none at all. A code's total silence as to a justification defense leaves it unclear whether or not the legislature meant for that justification to exist. It seems difficult to believe that none of the states receiving our lowest ranking wished their criminal law to impose liability for acts of self-defense. Yet this means that they expected such a justification to be imposed by judicial fiat. This is improper because it allows the courts to create any justifications they deem appropriate and ignore others that may well be appropriate. In other words, it allows for -- indeed, calls for -- legislation from the bench.

In contrast, the five highest-ranked codes all contain comprehensive general justification sections covering self-defense, defense of others, defense of property, acts undertaken within a law enforcement capacity, and acts by others having a special responsibility. Four of these five codes also contain a "choice of evils" or "necessity" justification provision. Probably the most comprehensive of all is New Jersey's, whose justification chapter contains nine substantive provisions as well as a definition section.

3. Definition of Terms

One function of a comprehensive criminal code is to provide notice to lay persons of the conduct it defines as criminal. The lowest-ranked codes under question one fail at that task. The criminal codes of North Carolina, South Carolina, Massachusetts, the District of Columbia, and Rhode Island do not contain general definition sections and rarely, if ever, contain definitions within specific sections. In fact, these codes define even serious offenses incompletely or not at all. For example, many of the lowest-ranked codes fail to define "assault." The Massachusetts code is typical; as to assault it simply says:

Whoever commits an assault or an assault and battery upon another shall be punished by imprisonment for not more than two and one half years in a house of Correction or by a fine of not more than five hundred dollars.

South Carolina's provision adds verbiage but not clarity or notice of the offense elements:

If any person be convicted of assault, assault and battery, assault or assault and battery with intent to kill or manslaughter and it shall appear upon the trial that the assault, assault and battery, assault or assault and battery with intent to kill or manslaughter shall have been committed with a deadly weapon of the character specified in § 16-23-460 carried concealed upon the person of the defendant so convicted the presiding judge shall, in addition to the punishment provided by law for such assault, assault and battery, assault or assault and battery with intent to kill or manslaughter, inflict further punishment upon the person so convicted by confinement in the Penitentiary for not less than three months nor more than twelve months, with or without hard labor, or a fine of not less than two hundred dollars or both fine and imprisonment, at the discretion of the judge.

Some codes fail to define even "murder" or "rape." These codes completely, and inappropriately, delegate the task of defining major offenses to the courts.

Even when offering some guidance as to the meaning of an offense, the worst codes frequently fail to provide definitions for specific terms they use in defining the offense. For example, Idaho's code uses the phrase "great bodily harm" to establish an aggravating factor for a variety of offenses, but nowhere does the code explain what constitutes "great bodily harm" or how it differs from simple "bodily injury." The District of Columbia's code similarly uses the phrase "serious bodily injury" to define offenses in chapter 5 (assault), but gives a meaning of that term only in chapter 41 (sexual abuse).

In contrast, the five highest-scoring codes are thorough in defining offenses. New Jersey's assault provision, both shorter and clearer than South Carolina's, offers an example:

A person is guilty of assault if he:

(1) Attempts to cause or purposely, knowingly or recklessly causes bodily injury to another; or

(2) Negligently causes bodily injury to another with a deadly weapon; or

(3) Attempts by physical menace to put another in fear of imminent serious bodily injury.

Such a concise formulation is possible because blanket terms -- such as "purposely," "knowingly," "recklessly," "negligently," and even "attempt" -- are defined in the code's general section. Additionally, the terms "bodily injury," "serious bodily injury," and "deadly weapon" -- which also recur in the code, but for a narrower range of offenses -- are defined in a separate definition section preceding the offense provisions for homicide, assault, kidnapping, and rape. This structure is typical of the best codes: providing a separate general "definitions" section, or a "definition" provision immediately preceding those offenses for which the definition is relevant, for recurring terms.

B. Effectiveness in Communicating the Rules of Conduct

The codes' scores on question 2 are as follows:

This question had an essentially bipolar distribution: a group of 33 codes receiving scores between 2.6 and 3.3, and a group of 14 codes receiving scores between 0.4 and 0.9. Such an outcome should hardly be unexpected, for this question weighs heavily the code's organization, and there are only a limited number of ways in which to organize a criminal code, making it difficult for many codes to distinguish themselves or receive deviating "outlier" scores. Most states categorized offenses by type. Nearly all of the others listed offenses alphabetically -- a less satisfactory organizational method. Additionally, few states stood out in terms of the clarity of their exposition of the rules of conduct, because even the best codes commonly use unnecessary legalese to set out the rules of conduct. Below we consider states' performances on each of our scoring factors.

1. Drafting Style: Generally

The five codes receiving the lowest scores for this question -- Rhode Island, West Virginia, Maryland, Massachusetts, and Mississippi -- all frequently use confusing, convoluted, or arcane language in setting out offenses. Consider, for example, Rhode Island's murder provision:

The unlawful killing of a human being with malice aforethought is murder. Every murder perpetrated by poison, lying in wait, or any other kind of willful, deliberate, malicious, and premeditated killing, or committed in the perpetration of, or attempt to perpetrate, any arson or any violation of §11-4-2, 11-4-3, or 11-4-4, rape, any degree of sexual assault or child molestation, burglary or breaking and entering, robbery, kidnapping, or committed during the course of the perpetration, or attempted perpetration, of felony manufacture, sale, delivery, or other distribution of a controlled substance otherwise prohibited by the provisions of chapter 28 of title 21, or while resisting arrest by, or under arrest of, any state trooper or police officer in the performance of his or her duty or committed against an assistant attorney general or special assistant attorney general in the performance of his or her duty; or perpetrated from a premeditated design unlawfully and maliciously to effect the death of any human being other than him or her who is killed is murder in the first degree. Any other murder is murder in the second degree. The degree of murder may be charged in the indictment or information, and the jury may find the degree of murder, whether the same be charged in the indictment or information or not, or may find the defendant guilty of a lesser offense than that charged in the indictment or information, in accordance with the provisions of § 12-17-14.

Who could seriously believe that this effectively communicates the rules of conduct to citizens, or to lawyers, for that matter? How "willful, deliberate, malicious, and premeditated killing" constitutes a discrete subset of killing "with malice aforethought" is unclear. The second sentence, comprised of 165 words, it resembles a maze designed for a lab rat.

The West Virginia statute banning desecration of the flag provides another example of tortuous sentence structure:

Any person who for exhibition or display shall place, or cause to be placed, any words, figures, marks, pictures, designs, drawings, or any advertisement of any nature, upon any flag, standard, color or ensign of the United States, or upon the state flag of this State, or shall expose or cause to be exposed to public view any such flag, standard, color or ensign, upon which shall have been printed, painted or otherwise placed, or to which shall be attached, appended, affixed or annexed, any words, figures, marks, pictures, designs, drawings, or any advertisement of any nature or kind, or who shall expose to public view, manufacture, sell, expose for sale, give away, or have in possession for sale or to give away, or for use for any purpose, any article or substance, being an article of merchandise or a receptacle of merchandise, or article or thing for carrying or transporting merchandise, upon which shall have been printed, painted, attached or otherwise placed, a representation of any such flag, standard, color or ensign, to advertise, call attention to, decorate, mark or distinguish the article or substance on which so placed, or who shall publicly mutilate, deface, defile or defy, trample upon or cast contempt, either by words or acts, upon any such flag, standard, color or ensign, he shall be deemed guilty of a misdemeanor, and, upon conviction, shall be fined not less than five nor more than one hundred dollars, and may, at the discretion of the court or justice [magistrate] trying the case, be confined in jail for a period not exceeding thirty days. Any justice of the peace [magistrate] of the county wherein the offense was committed shall have concurrent jurisdiction of such offense with the circuit or other courts of such county. The words "flag, standard, color or ensign of the United States," as used in this section, shall be construed to include any flag, standard, color, ensign, or any representation or picture of a flag, standard, color or ensign, made of or upon any substance whatever, and of any size whatever, showing the national colors, the stars and stripes. This section shall not apply to any act permitted by the statutes of the United States, or of this State, or by the regulations of the United States army and navy, or of the national guard of this State, or of the members of the department of public safety; nor shall this section be construed to apply to the regular issue of a newspaper or other periodical, or to any book, certificate, diploma, warrant or commission, on which shall be printed said flag, disconnected from any advertisement, or to the vignette of any political ballot.

The presence of multiple terms and confusing sentences detracts from the comprehensibility of the provision. The passage uses long multi-clause sentences with no organizational mechanism, such as numbering or lettering, to break down and readily identify when and where a reader should pause and consider what she just read as a distinct unit of information.

By contrast, the highest-scoring criminal codes are more clearly written and easily understood. They use clearer language and easier-to-read sentences, and sections are short enough to be able to read through (as contrasted with the sections outlined above). For example, compare Rhode Island's opaque murder provision with these provisions from the Hawaii code:

§ 707-701 Murder in the first degree

(a) A person commits the offense of murder in the first degree if the person intentionally or knowingly causes the death of:

(b) More than one person in the same or separate incident;

(c) A peace officer, judge, or prosecutor arising out of the performance of official duties,

(d) A person known by the defendant to be a witness in a criminal prosecution;

(e) A person by a hired killer, in which event both the person hired and the person responsible for hiring the killer shall be punished under this section; or

(f) A person while the defendant was imprisoned.

(g) Murder in the first degree is a felony for which the defendant shall be sentenced to imprisonment as provided in section 706-656. . . .

§ 707-701.5 Murder in the second degree

(a) Except as provided in section 707-701, a person commits the offense of murder in the second degree if the person intentionally or knowingly causes the death of another person.

(b) Murder in the second degree is a felony for which the defendant shall be sentenced to imprisonment as provided in section 706-656.

Although Hawaii's provision contains at least as much information and refinement in specifying types of prohibited conduct as Rhode Island's, it lays out that information in a more straightforward way. The above provision uses everyday parlance instead of overly legalistic terms like "deliberately premeditated malice aforethought." Multi-clause sentences are replaced with either short and concise sentences or are broken down and organized through numbering or lettering.

2. Drafting Style: Simplicity of Justification Rules

With respect to the provision of justification rules, a low-scoring code may err in either direction: some codes fail entirely to provide any justification rules; others provide rules that are so complex as to make their practical application impossible, hence -- in terms of notice and behavior modification -- nearly as bad as providing no rule at all. In the latter category fall such states as Nebraska and Connecticut. Nebraska's self-defense provision reads as follows:

Use of force in self-protection.

(1) Subject to the provisions of this section and of section 28-1414, the use of force upon or toward another person is justifiable when the actor believes that such force is immediately necessary for the purpose of protecting himself against the use of unlawful force by such other person on the present occasion.

(2) The use of such force is not justifiable under this section to resist an arrest which the actor knows is being made by a peace officer, although the arrest is unlawful.

(3) The use of such force is not justifiable under this section to resist force used by the occupier or possessor of property or by another person on his behalf, where the actor knows that the person using the force is doing so under a claim of right to protect the property, except that this limitation shall not apply if:

(a) The actor is a public officer acting in the performance of his duties or a person lawfully assisting him therein or a person making or assisting in a lawful arrest;

(b) The actor has been unlawfully dispossessed of the property and is making a reentry or recapture justified by section 28-1411; or

(c) The actor believes that such force is necessary to protect himself against death or serious bodily harm.

(4) The use of deadly force shall not be justifiable under this section unless the actor believes that such force is necessary to protect himself against death, serious bodily harm, kidnapping or sexual intercourse compelled by force or threat, nor is it justifiable if:

(a) The actor, with the purpose of causing death or serious bodily harm, provoked the use of force against himself in the same encounter; or

(b) The actor knows that he can avoid the necessity of using such force with complete safety by retreating or by surrendering possession of a thing to a person asserting a claim of right thereto or by complying with a demand that he abstain from any action which he has no duty to take, except that:

(i) The actor shall not be obliged to retreat from his dwelling or place of work, unless he was the initial aggressor or is assailed in his place of work by another person whose place of work the actor knows it to be; and

(ii) A public officer justified in using force in the performance of his duties or a person justified in using force in his assistance or a person justified in using force in making an arrest or preventing an escape shall not be obliged to desist from efforts to perform such duty, effect such arrest or prevent such escape because of resistance or threatened resistance by or on behalf of the person against whom such action is directed.

(5) Except as required by subsections (3) and (4) of this section, a person employing protective force may estimate the necessity thereof under the circumstances as he believes them to be when the force is used, without retreating, surrendering possession, doing any other act which he has no legal duty to do, or abstaining from any lawful action.

(6) The justification afforded by this section extends to the use of confinement as protective force only if the actor takes all reasonable measures to terminate the confinement as soon as he knows that he safely can do so, unless the person confined has been arrested on a charge of crime.

Connecticut's corresponding provision is only slightly less verbose. It seems unrealistic to think that such provisions could give real guidance to a person caught in a self-defense situation.

The better codes, on the other hand, contain justification rules that are easier for a lay person to understand and remember. An example is Oregon's provisions covering the defense of oneself or another person, which read as follows:

161.209. Use of physical force in defense of a person.

Except as provided in ORS 161.215 and 161.219, a person is justified in using physical force upon another person for self-defense or to defend a third person from what the person reasonably believes to be the use or imminent use of unlawful physical force, and the person may use a degree of force which the person reasonably believes to be necessary for the purpose.

161.215. Limitations on use of physical force in defense of a person.

Notwithstanding ORS 161.209, a person is not justified in using physical force upon another person if:

(1) With intent to cause physical injury or death to another person, the person provokes the use of unlawful physical force by that person; or

(2) The person is the initial aggressor, except that the use of physical force upon another person under such circumstances is justifiable if the person withdraws from the encounter and effectively communicates to the other person the intent to do so, but the latter nevertheless continues or threatens to continue the use of unlawful physical force; or

(3) The physical force involved is the product of a combat by agreement not specifically authorized by law.

161.219. Limitations on use of deadly physical force in defense of a person.

Notwithstanding the provisions of ORS 161.209, a person is not justified in using deadly physical force upon another person unless the person reasonably believes that the other person is:

(1) Committing or attempting to commit a felony involving the use or threatened imminent use of physical force against a person; or

(2) Committing or attempting to commit a burglary in a dwelling; or

(3) Using or about to use unlawful deadly physical force against a person

Though hardly simple, this provision at least holds open the possibility that an average citizen could remember the justification rule and apply it properly, as opposed to the complexity found in the Nebraska law, which would require one to obtain a legal consultation -- and an expensive one, at that -- before feeling able to defend oneself with legal impunity.

3. Organization: Generally

The lowest-scoring codes make clear the pitfalls of alphabetical, or completely random, organization. Not only does Vermont's code organize chapters alphabetically, but it often places offenses into chapters almost arbitrarily: robbery appears in the chapter on assault, while assault appears in the chapter on "breaches of the peace"! Mississippi groups offenses into chapters by category, but organizes within chapters alphabetically, thereby creating organizational confusion within chapters. The first three provisions of its chapter on "crimes versus the person" are as follows: "abduction for purposes of marriage"; "abortion; advertisement, sale or gift of drugs or instruments" (one section containing two different kinds of offenses!); and "simple assault; aggravated assault." The "rape" offenses in that chapter are separated from the "sexual battery" offenses by robbery, extortion, robbery again (!), "threats and intimidation," and three provisions on "timber, trees and saw logs." At the end of the chapter comes the "Mississippi Carjacking Act," an obvious later addition that the legislature did not bother to fit into the alphabetical scheme. Why "extortion" would be placed between offenses starting with the letter "r," or why "sexual assault" would follow offenses starting with the letter "t," is unknown. What is obvious, though, is the confusion likely to result from an organizational system (if it can be called such) that widely separates "rape" and "sexual battery," which most people probably would not even suspect to constitute separate categories of crime.

Some codes not only fail to provide offenses in order of seriousness, but rather use no apparent order at all. Massachusetts lists "indecent assault and battery upon persons 14 and older" before the more serious "assault and battery on a child" and "assault and battery upon an elderly or disabled person," all of which are followed by the still more serious crime of "mayhem," which is followed by the less serious crime of "assault; intent to murder or maim," which is followed by the more serious "dangerous weapon; assault in dwelling house."

The best-organized codes all reflect the indelible influence of the Model Penal Code. Indeed, general organizational structure may be the best available example of a useful organizational innovation offered by the Model Penal Code and emulated by a large number of state criminal codes. The general organizational scheme is easy to recognize: a distinct "general part" (containing principles of liability, justifications, responsibility, and inchoate crimes) followed by a "special part" grouping offenses into categories (Offenses Involving Danger to the Person, Offenses Against Property, Offenses Against the Family, Offenses Against Public Administration, and Offenses Against Public Order and Decency). Additionally, offenses within each category are arranged in decreasing order of seriousness. The organizational schemes of the codes of Colorado, Hawaii, Alaska, New Jersey, and several other states generally reflect the structure of the Model Penal Code.

4. Organization: Overlapping Offenses

Codes that scored well on question 2 generally avoided the existence of multiple offenses governing the same criminal conduct. Nonetheless, top-scoring states such as Colorado, Alabama, Hawaii, and Minnesota also contain general provisions expressly governing the prosecution of offenses whose definitions overlap. These codes typically are emulating the Model Penal Code's provisions on the subject.

Few of the lowest-scoring codes, and none of the bottom five, offer any guidance regarding how to deal with overlapping offenses, even though examples of overlapping offenses abound throughout those codes. For example, Mississippi has numerous code sections relating to homicide, as even a cursory review of section headings reveals.

§ 97-3-15. Homicide; justifiable homicide.

§ 97-3-17. Homicide; excusable homicide.

§ 97-3-19. Homicide; murder defined; capital murder.

§ 97-3-21. Homicide; penalty for murder or capital murder.

§ 97-3-23. Homicide; death following duels fought out of state.

§ 97-3-25. Homicide; penalty for manslaughter.

§ 97-3-27. Homicide; killing while committing felony.

§ 97-3-29. Homicide; killing while committing a misdemeanor.

§ 97-3-31. Homicide; killing unnecessarily, while resisting effort of slain to commit felony or do unlawful act.

§ 97-3-33. Killing trespasser involuntarily.

§ 97-3-35. Homicide; killing without malice in the heat of passion.

§ 97-3-37. Homicide; killing of an unborn quick child.

§ 97-3-39. Homicide; drunken doctor, etc., unintentionally causing death.

§ 97-3-41. Homicide; overloading boat.

§ 97-3-43. Homicide; ignorant or negligent management of steamboat or railroad engine.

§ 97-3-45. Homicide; owner of dangerous animal.

§ 97-3-47. Homicide; all other killings.

§ 97-3-49. Suicide; aiding.

§ 97-3-109. Drive-by shooting; drive-by bombing.

While it may add precision to distinguish homicide by overloading a boat from homicide by negligent management of a steamboat -- both of which doubtless occur frequently enough to merit separate homicide provisions -- it unnecessarily complicates the code, making it more difficult to determine under which section certain conduct may fall or to determine why certain conduct would not be covered by the more general homicide provisions.

Other instances of overlapping coverage of violent conduct include Massachusetts' inclusion of separate offenses banning prize fighting and boxing matches and Maryland's overlapping sections for manslaughter and manslaughter by automobile or vessel. Notwithstanding its three general provisions for willful trespass, Michigan's code also adds specific provisions for trespass on cranberry marshes; huckleberry and blackberry marshes; vineyards, orchards or gardens; or trespass to destroy or remove specified "medicinal plants." The text of the provisions makes clear that the specific provisions are mere surplusage. Consider, for example:

Sec. 547. Wilful Trespass by Entering Improved Land of Another With Intent to Injure or Destroy -- Any person who shall wilfully commit any trespass by entering upon the garden, orchard or other improved land of another, without permission of the owner thereof, and with intent to cut, take, carry away, destroy or injure the trees, grain, grass, hay, fruit or vegetables there growing or being, shall be guilty of a misdemeanor. . . .

Sec. 550. Trespass upon Vineyards, Orchards or Gardens -- Any person who shall enter a vineyard, orchard or garden, without the consent of the owner, and pick, take, carry away, destroy or injure any of the fruits, vegetables or crops therein, or in anywise injure or destroy any bush, tree, vine or plant, shall be guilty of a misdemeanor.

The two offenses are not exactly congruent -- the latter provision covers the act of "taking" as well as the intent -- but this is more confusing than enlightening, especially since the same punishment is provided in both cases. A single provision covering all relevant conduct would be clearer and avoid unnecessary questions of statutory interpretation.

C. Comprehensiveness and Accessibility of the Principles of Adjudication

The codes' scores on question 3 are as follows:

Perhaps the most remarkable aspect of the scoring results for question 3 is the wide discrepancy between the states that performed the best and those that performed the worst. The average scores for the top quartile (3.573) and top half (3.394) were the highest for any question, edging out the scores for question 1. On the other hand, fully six states received scores of 0 on this question. It is not especially hard to fathom the reasons for this result. Question 3 is similar to question 1 in that it addresses a fundamental element of a criminal code: the inclusion of rules governing adjudication of criminal offenses. But because the issues relating to question 3 are more subtle -- relating not to the obvious behavioral question of what acts are crimes, but the more sophisticated moral question of what actors are criminals -- some states either fail to recognize the need to address those issues or, perhaps worse, deliberately abrogate that responsibility, leaving it to the discretion of individual courts.

Probably the best example of wide disparity in a code's approach to a comprehensive adjudication code is the inclusion of explicit culpability provisions. Most of the states that include such provisions follow the language of the Model Penal Code, which is very detailed. Many other states, however, fail to codify any culpability term definitions and therefore receive low scores (or null scores). Below we consider states' performances on each of our scoring factors.

1. Specification of Necessary Rules: Culpability requirements

A common characteristic of codes that scored poorly on question 3 is the failure to define culpability terms and terms used within the code's adjudication provisions. Consider the example of manslaughter, an offense whose conduct component -- killing a human being -- is identical to that for the offense of murder; the two differ only with respect to the requisite level of culpability on the offender's part. The following is the full extent of North Carolina's manslaughter provisions:

Voluntary manslaughter shall be punishable as a Class D felony, and involuntary manslaughter shall be punishable as a Class F felony.

This provision offers no guidance regarding the meaning of the term "manslaughter" or how the two forms of manslaughter differ from one another or from murder. Alternatively, some other states at the bottom of the ranking use culpability terms but do not explain their meaning. For example, Wyoming's manslaughter provision contains but does not define the term "heat of passion." This is almost as bad as not using culpability terms at all, since these terms are opaque on their face and must be clarified by common law decisions.

Many of these same states also provide poorly defined or very loose culpability requirements for murder. South Carolina's murder provision reads:

"Murder" is the killing of any person with malice aforethought, either express or implied.

Neither here nor anywhere else does the code define "malice aforethought" (either express or implied), the culpability requirement of the offense. Similarly, Wyoming's murder provision contains the term "premeditated malice" but does not define it.

Those states that are thorough in specifying culpability requirements tend to emulate, if not parrot, the Model Penal Code. These codes define, in their general parts, a limited set of terms (almost always "intentionally" or "purposely," "knowingly," "recklessly," and "negligently") that are then used throughout the special part of the code to set out a culpability requirement for each element of each offense. New Jersey, Colorado, North Dakota, Missouri, Illinois, Texas, Hawaii, and a number of other high-scoring states follow the Model Penal Code in this respect.

2. Specification of Necessary Rules: Excuses and Nonexculpatory Defenses

North Carolina, Michigan, Massachusetts, West Virginia, Rhode Island, Mississippi, and Maryland are among the states that fail to define any excuses or nonexculpatory defenses in their penal codes. Numerous other codes include only a fraction of the commonly recognized excuses and nonexculpatory defenses. For example, Iowa codifies only duress and insanity, and New Mexico's penal code contains only a statute-of-limitations provision.

But even where excuses and nonexculpatory defenses are codify, the provisions often are of little help when inadequately defined. Some codes grant excuses to broad categories of actors, such as "the insane," without explaining how the adjudicator is to determine who fits into such a category. Oklahoma's penal code is illustrative:

Persons capable of committing crimes--Exceptions--Children--Idiots--Lunatics--Ignorance--Commission without consciousness--Involuntary subjection.

All persons are capable of committing crimes, except those belonging to the following classes:

1. Children under the age of seven (7) years.

2. Children over the age of seven (7) years, but under the age of fourteen (14) years, in the absence of proof that at the time of committing the act or neglect charged against them, they knew its wrongfulness.

3. Idiots.

4. Lunatics, insane persons, and all persons of unsound mind, including persons temporarily or partially deprived of reason, upon proof that at the time of committing the act charged against them they were incapable of knowing its wrongfulness.

5. Persons who committed the act, or made the omission charged, under an ignorance or mistake of fact which disproves any criminal intent. But ignorance of the law does not excuse from punishment for its violation.

6. Persons who committed the act charged without being conscious thereof.

7. Persons who committed the act, or make the omission charged, while under involuntary subjection to the power of superiors.

It appears that in Oklahoma, the fact that a criminal is an idiot might actually be his saving grace. California has a similar provision.

By contrast, the codes that received the highest scores for question 3 codified a wide range of excuses and nonexculpatory defenses. For example, New Jersey codifies defenses for mistakes of fact or law, involuntary intoxication, duress, lack of consent, entrapment, insanity, immaturity, and involuntary acts, and provides a statute of limitations. Illinois, Missouri, Hawaii, Colorado, and Texas are comparable.

3. Specification of Necessary Rules: Other Adjudication Issues

The codes' performance with respect to the inclusion of general adjudicatory provisions, covering issues such as causation, complicity, consent, and multiple offenses, follows a common pattern whereby the codes may be divided into three groups. In the first group are the codes that provide no rules at all to govern these issues. The second group includes codes that offer provisions addressing the basics of these issues. The third group of states offers provisions that are more comprehensive, dealing with not only the basics but also a variety of related issues. For example, Hawaii's code deals not only with the basic "but for" cause requirement but also issues that lawyers would call "proximate cause" and "transferred intent":

Causal relationship between conduct and result.

Conduct is the cause of a result when it is an antecedent but for which the result in question would not have occurred.

Intentional or knowing causation; different result from that intended or contemplated.

In the following instances intentionally or knowingly causing a particular result shall be deemed to be established even though the actual result caused by the defendant may not have been within the defendant's intention or contemplation:

(1) The actual result differs from that intended or contemplated, as the case may be, only in the respect that a different person or different property is injured or affected or that the injury or harm intended or contemplated would have been more serious or more extensive than that caused; or

(2) The actual result involves the same kind of injury or harm as the intended or contemplated result and is not too remote or accidental in its occurrence or too dependent on another's volitional conduct to have a bearing on the defendant's liability or on the gravity of the defendant's offense.

Hawaii's code also includes similarly thorough provisions covering complicity, consent, multiple convictions, and liability of corporations. The codes of New Jersey, Colorado, and North Dakota are also notable for their detailed specification of general adjudicatory provisions.

4. Sufficiency of Detail and Ease of Use

Of course, since many states at the bottom proved to have no adjudication provisions whatever, it is impossible to analyze whether they go into sufficient detail. The weakness of such undefined terms as South Carolina's "malice aforethought" or Wyoming's "premeditated malice," noted above, is also obvious. The term "malice" is inscrutable on its face and requires elaboration for adjudicators properly to distinguish between actors who are suitably blameworthy and those who are not. Such broad and cryptic culpability terms suffer greatly by comparison with the definitions of terms such as "purpose" under state code provisions, such as this New Jersey provision, that follow the Model Penal Code's lead:

A person acts purposely with respect to the nature of his conduct or a result thereof if it is his conscious object to engage in conduct of that nature or to cause such a result. A person acts purposely with respect to attendant circumstances if he is aware of the existence of such circumstances or he believes or hopes that they exist.

Even when the term "malice" is given a definition, the state it describes often refers to a complex relation between thought, emotion, moral sentiment, and deliberative intent. The above definition of "purpose," on the other hand, focuses on one's immediate attitude toward one's conduct, a benchmark that is both more appropriate for adjudicative determinations and more objectively demonstrable than the notion of "malice."

The same disparity between broad, amorphous terms and detailed definitions exists with respect to defenses. As we have seen, it is an absolute defense to criminal liability in California to be an "idiot." Contrast that with Hawaii's excuse provision, again based on the Model Penal Code, covering "mental defect":

A person is not responsible, under this Code, for conduct if at the time of the conduct as a result of physical or mental disease, disorder, or defect the person lacks substantial capacity either to appreciate the wrongfulness of the person's conduct or to conform the person's conduct to the requirements of law.

This provision, and others like it, directly describe the relation of intellectual impediment to legal responsibility, the proper enterprise for an adjudicative determination. The definition is subjective and calls for a complex intuitive judgement by the decisionmaker, but that is inevitable in the formulation of an excuse defense that tries to capture our complex notions of blameworthiness. The Hawaii formulation at least gives the decisionmaker a decisional framework for considering the question, something that the "idiot" formulation does not even hint at.

D. Accuracy in Imposing Liability

The codes' scores on question 4 are as follows:

In terms of overall average score, the codes performed worst on this question: the 1.803 mean was the lowest for any question by more than 0.3 points. Only twelve states received scores above 3.0, fewer than for any question except question 5. Even more glaring are the low-end scores: seventeen codes received scores of 0.5 or lower, and the average score for the lowest quartile is a microscopic 0.096. This reflects a combination of the two trends we note elsewhere; that codes receive lower scores for our adjudication criteria than for our conduct criteria, and lower scores on "quality" measures than on "comprehensive" measures. Many states simply offer so little guidance, or such haphazard guidance, as to liability determinations that they are scarcely, if at all, superior to outright delegation of those determinations to the courts. Below we consider states' performances on each of our scoring factors.

1. Appropriate Criminalization Decisions

The most obvious and probably most plentiful, if not the most significant, flaw of low-scoring criminal codes is their criminalization of conduct that is harmless at best. The codes of North Carolina, West Virginia, Mississippi, Maryland, Michigan, and a number of other low scorers contain numerous provisions prohibiting morally and practically trivial activities. As if conceding the arbitrariness of these determinations as to what acts the state should sanction through criminal punishment, a few codes include offenses that apply only within limited parts of their area of jurisdiction, as with Maryland's law against fortune-telling:

In Caroline County, Carroll County, and in Talbot County every person who shall demand or accept any remuneration or gratuity for forecasting or foretelling or for pretending to forecast or foretell the future of another by cards, palm reading or any other scheme, practice or device, shall be deemed guilty of a misdemeanor and upon conviction thereof shall be punished by a fine of not more than $100.00 or be imprisoned in the county jail for a term of not more than six months, or both, in the discretion of the court.

Are the citizens of other counties unworthy of protection from fortune-telling charlatans, or are the citizens of the three named counties considered unusually susceptible to such ruses? However one reads this provision, some of the people of Maryland have good reason to complain about the reach of this law -- or would, were it not so utterly silly in the first place. Maryland's code also includes no fewer than three provisions -- each dealing with separate counties, using distinct wording to define the offense, and providing for distinct punishment! -- for prosecuting dog owners who fail to confine dogs in heat.

In a similar vein, Michigan appears to have more concern about the behavior of sleigh-drivers on its upper than on its lower peninsula:

Bells on Sleighs and Cutters in Upper Peninsula -- Any person who shall drive or cause to be driven, faster than a walk, on any public highway or private road used by the public in this state, or on any street of an incorporated city or village thereof any sleigh or cutter or other vehicle used as a substitute for either, drawn by horses or mules, or by horse or mule, during the season of sleighing without having bells on at least 1 of the animals so used or without having bells attached to such sleigh, cutter or other vehicle so drawn, in such a manner as to warn foot travelers of its approach, he or they shall be guilty of a misdemeanor: Provided, That the provisions of this section shall apply to the upper peninsula alone.

Many of Michigan's other offenses, though at least uniform in their application throughout the state, are no less absurd. Michigan makes it a crime to use "any indecent, immoral, obscene, vulgar or insulting language in the presence or hearing of any woman or child," and provides for up to five years' imprisonment for "[a]ny man who shall seduce and debauch any unmarried woman." Michigan also devotes an entire chapter of its penal code to the prohibition of performing the national anthem with "embellishments of national or other melodies," or "as a part or selection of a medley of any kind," or "for dancing or as an exit march." At least one of this Article's co-authors, and doubtless some readers who concern themselves with criminal law, may well be subject to prosecution in Michigan for publication of any printed material "principally made up of criminal news, police reports or accounts of criminal deeds or pictures, stories of deeds of bloodshed, lust or crime."

Michigan hardly has a monopoly on criminal minutiae, however. North Carolina makes a criminal of "[a]ny person who bribes, or offers to bribe, any judge or other official in any horse show, with intent to influence his decision or judgment concerning said horse show." California is more concerned with horses themselves, prohibiting anyone from deliberately tripping or poling one of them. It is fortunate for Julius Erving that no NBA franchise exists in West Virginia, as his conducting business there under his usual basketball sobriquet would be punishable by the state:

Unlawful use of prefix "Doctor" or "Dr."; penalty.

It shall be unlawful for any person to use the prefix "Doctor" or "Dr." in connection with his name in any letter, business card, advertisement, sign or public display of any nature whatsoever, without affixing thereto suitable words or letters designating the degree which he holds. Any person who shall violate the provisions of this section shall be guilty of a misdemeanor, and upon conviction thereof shall be fined for each such offense not less than ten nor more than five hundred dollars, or imprisoned in the county jail not more than twelve months, or both fined and imprisoned, in the discretion of the court.

A number of states contain archaic provisions prohibiting duels or challenges to duel; Arkansas and some other states also make it a felony to "proclaim any other person as a coward, or use any other opprobrious or abusive language, for not accepting a challenge to fight a duel or not fighting a duel." These examples of trivial criminal offenses or offenses better dealt with elsewhere are only a small sampling of such provisions in lower scoring states. Certain states just seem to have a knack for criminalizing the trivial.

Our better scoring codes do not criminalize trivial behavior. Rather, they concern themselves more with behavior that has a clearly identifiable victim or that disrupts the proper and fair operation of the government, than with identifying "offenses" more properly remediable through private lawsuits or noncriminal regulation. This is not to say that lapses in judgment by these legislatures do not occur. For high-scoring codes, though, such incidents are rare and are not exacerbated by limiting the application of trivial offenses to certain geographic regions of the state (which seems to be a recurring theme in the case of the lower-ranking codes).

Moreover, the higher-scoring states presumably recognize that the kind of behavior typically criminalized by such specialized provisions, if it indeed merits criminal sanction, will frequently already be penalizable under another, more general provision of the criminal code. For example, a violation of Michigan's provision regarding "Bells on Sleighs and Cutters in the Upper Peninsula" would presumably also violate a typical reckless endangerment statute -- and if it would not, such behavior arguably would therefore not constitute a sufficient danger to merit criminal punishment.

Another criterion for gauging how accurately a criminal code metes out criminal liability is whether the code establishes satisfactorily demanding conduct requirements for inchoate offenses such as attempt and conspiracy. The lower-ranked codes either do not provide a general definition of attempt or provide that "any act" toward the commission of a crime constitutes an attempt. Mississippi's definition of attempt is typical in this regard, providing for attempt liability for "[e]very person who shall design and endeavor to commit an offense, and shall do any overt act toward the commission thereof, but shall fail therein, or shall be prevented from committing the same."

A handful of states, however, do impose meatier conduct requirements for prosecution of attempted offenses. Texas, for instance, defines an attempt as "an act amounting to more than mere preparation that tends but fails to effect the commission of the offense intended." Ohio, meanwhile, makes clear the importance of intervening events as the only thing separating its definition of attempt from an otherwise enumerated crime.

2. Appropriate Liability Rules

In addition to including rules of conduct that outlaw behavior for which criminal sanction is inappropriate, low-ranking states include rules of adjudication that fail to protect from criminal liability conduct that lacks sufficient blameworthiness. An important provision in this respect is a general "read-in" rule specifying the default culpability requirement that will apply when an offense provision is silent as to the culpability requirement. The worst codes, as noted earlier, fail to include any general part that delineated culpability requirements in the first place. Several states with middle-of-the-road scores include culpability rules, but have no read-in provision.

The highest scoring states under question 4 -- such as Hawaii, New Jersey, Texas, Arkansas, and Alaska -- all contain read-in provisions. Most states provide for recklessness as the default culpability requirement for any element of an offense for which the requirement is unspecified. A notable exception is Alaska, which is even more particular about its read-in provision, stating that absent a specific prescription, "the culpable mental state that must be proved with respect to (1) conduct is 'knowingly'; and (2) a circumstance or a result is 'recklessly.'"

Some states laudably attempt to provide a read-in provision, but they significantly expand the scope and application of various offenses in an unjust way by providing for negligence as the default culpability requirement. With negligence as a baseline culpability requirement, a legislature's failure to specify the culpability requirements for an offense element may have significant consequences. Oregon is one state with such a read-in provision.

Another example of a failure to limit liability to the degree of a defendant's blameworthiness is an "unconstrained" felony-murder rule, i.e., a rule imposing liability for murder for any death that results during the commission of any felony, regardless of the intention of the "murderer." New Mexico, for instance, includes as first-degree murder "the killing of one human being by another without lawful justification or excuse . . . in the commission of or attempt to commit any felony." Virginia includes unconstrained felony murder as second-degree murder. In addition to a felony-murder rule for specified felonies and a residual provision categorizing any homicide during any other felony as manslaughter, Mississippi provides for a "misdemeanor manslaughter" rule. Another common inappropriate liability rule found in low-scoring codes allows juries to render a verdict of "guilty but mentally ill."

The best codes, although usually codifying some form of felony-murder rule, enumerate specific felonies to which the rule will apply. Even better are states such as New Jersey and North Dakota, whose codes, in addition to limiting the felonies to which the rule applies, limit its potential to impose guilt without blameworthiness by creating "affirmative defenses" for situations tending to negate culpability. This approach is functionally similar to the Model Penal Code's rule, which no state has adopted wholesale, that death caused during the commission of a felony creates a (rebuttable) "presumption" of culpability. Only Hawaii has no felony-murder rule at all. None of the best codes, however, attempt to circumvent the insanity defense though a "guilty but mentally ill" verdict.

3. A Comprehensive System of Defenses

The lowest-scoring states under question 4 codify no excuse defenses or justification defenses whatsoever. For the high-ranking states, it is clear that the Model Penal Code has set the standard of performance. The excuse defenses of the best states very nearly mirror those provided in the Code, although no state includes every Model Penal Code defense. For example, Hawaii tracks the Model Penal Code's excuse provisions -- including involuntary act, ignorance or mistake, intoxication, duress, military orders, consent, de minimis infractions, entrapment, and insanity -- with the exception of the defense of immaturity. New Jersey has no "military orders" defense, but was given extra credit as the only state that provided a true "reasonable mistake of law" defense. These states also follow the Model Penal Code's lead with respect to justification defenses.

A few states scored well under question 4 despite paltry showings in the excuse defense department. These codes managed to score well because they are comprehensive in other areas -- such as read-in provisions and justification defenses, and a lack of trivial offenses -- and because the excuses they do include are the more significant than those they omit. For example, Texas' lack of military orders, consent, and de minimis defenses is counterbalanced by that code's provision of mistake, insanity, involuntary intoxication, duress, entrapment, and immaturity defenses in its code. Additionally, the provision of some excuse and nonexculpatory defenses at least takes out of the purview of the courts the option as to whether or not to allow such defenses. Nonetheless, the more thorough a code is in providing excuse and nonexculpatory defenses, the greater our faith in its reliability in accurately assessing criminal liability.

E. Accuracy in Grading Liability and Punishment

The codes' scores on question 5 are as follows:

In contrast to the scoring for the other criteria, few states scored extremely well or extremely poorly on this question: this was the only question for which no state scored above 3.5 and no state scored below 0.5. Also, unlike the question 2 scoring pattern, the distribution of scores is even rather than clustered. For example, thirty-six codes received scores in the middle range between 1.0 and 3.0 (and about half of those were in the middle half, 1.5-to-2.5, of that range), by far the most for any question. It is perhaps especially difficult for codes to obtain a very high score for this criterion because of the difficulty in attaining a truly comprehensive, consistent, and rational grading system. It is quite possible for a code to be completely comprehensive in elaborating rules, for example, requiring no further effort in that area, but it is less likely that one will conclude that a state cannot improve on its grading scheme, either by making more gradations of punishment level or finer distinctions between offenses. On the other hand, no code utterly fails to satisfy the dictates of this criterion, because every code offers some suggestion of the appropriate punishment level for a criminal offense, and any limitation on punishment that a code provides is better than none. Only a state that provided no grading recommendations whatever, or actually curtailed adjudicators' discretion in a harmful way, could be as bad as total abdication of the grading function (which would merit a score of zero). Even the worst states at least take seriously the need to provide maxima for punishment, and that offers some utility. Below we consider states' performances on each of our scoring factors.

1. Consistency and "Grading" of Offenses

Perhaps the easiest way to get at the issue of consistency -- or rather, inconsistency, as this method will turn up only bad examples -- is to see whether the state punishes apparently similar offenses very differently. One remarkable example is Massachusetts, which establishes a maximum sentence of three months for participation in a "boxing match," but a maximum sentence of ten years for the crime of "prize fighting." Because neither of these terms is defined, it is not clear from the statute's face whether there is any difference (a paying audience? an award to the winner?) between the types of conduct proscribed by these two offenses. Whatever distinction may exist, however, surely it does not merit a four-thousand-percent increase in potential punishment as a crime escalates from the relatively innocuous practice of a "boxing match" to the grisly barbarity of a "prize

fight." Even Louisiana, whose code is hardly a model of technical efficiency, manages to include all illegal boxing competitions in a single provision with a single specified punishment range.

On the other hand, it is difficult, if indeed it makes conceptual sense at all, to list or briefly describe specific good examples of consistency within a code, since this criterion addresses each code's scheme of punishment as a whole. For this reason, and also because a code's relative degree of satisfaction of the consistency standard frequently can be recognized only in the breach, there are few "positive role models" that can be summarized briefly.

That said, some codes offer certain structural features that make clear, at the very least, that the legislature has taken seriously the project of grading offenses consistently. Alaska, for example, does not merely set out the sentences that attach to each grade of crime, but provides general descriptions of the types of crimes that should fall into each grade. South Carolina, after setting out its grading categories, lists all of the offenses that fall into each category, enabling immediate comparison. Unfortunately, these codes' substantive grading decisions did not always parallel their laudable efforts to make transparent (in one case) the guiding principles driving the grading of offenses, and (in the other) the results following from grading; neither of these two states ultimately fared especially well under our scoring system. Still, Alaska received a respectable score of 2.9, and South Carolina performed better on question 5 than on any other question, even though its treatment of inchoate offenses was among the worst of any state.

In addition to direct evidence from legislative decisions about punishments for specific offenses, a code's potential to achieve consistency derives from the nature of the code's "grading" scheme, if any. Arizona, for instance, includes six classes of felonies, three classes of misdemeanors, and a category of "petty offenses." Other states are similarly sophisticated. At the other extreme, some states lack any grading system whatever. On the other hand, the mere existence of a grading system would not completely shield a state from a score reduction if it appeared that the state did not take the grading project seriously. For example, Virginia's criminal code includes an elaborate grading system (six felony categories, four misdemeanor categories, and a separate category for non-offense "violations"), but includes many offenses for which a specific sentence is provided rather than an offense grade. This system obviously undercuts the utility of the grading system as a means of ensuring uniformity and consistency in punishment.

2. Recognizing Appropriate Distinctions

In addition to grading similar offenses similarly, a code must recognize relevant differences in culpability level, conduct, or results that distinguish different "degrees" of what might otherwise be the same criminal offense. Some of these distinctions are general and apply across various offenses. The distinction between inchoate and completed offenses is an example of such a broad-based distinction. Many states fail to consider the presence or absence of the harm resulting from a completed offense to be significant in determining the proper sentence. Some explicitly state that attempts are to be punished similarly to completed offenses: the South Carolina criminal code, for example, insists that "[a] person who commits [an] attempt, upon conviction, must be punished as for the principal offense." In a unique articulation, Mississippi quite charitably provides that an attempt may not be punished more than the principal offense. Some other states are silent as to the punishment of attempts vis-a-vis completed offenses generally, but proceed to define offenses to include attempt on the same footing as the completed act, thus allowing (if not ensuring) that the two will be treated equally in terms of punishment.

Other, more specific distinctions constitute relevant aggravating and mitigating factors that apply to particular offenses. Reviewing the codes' treatments of such factors can make one feel like Goldilocks: some states recognize too many factors, establishing different penalties based on ultimately irrelevant concerns; some recognize too few, making only the crudest efforts to acknowledge factors favoring an increase or decrease in punishment; a scant minority of states manage to get it just right. This section will present examples of the codes' best and worst efforts to recognize such factors with respect to three categories of offense: assault, arson, and theft.

Three factors seem most relevant in distinguishing the relative severity of assaults: the harm the assailant inflicted on the victim; the harm the assailant intended to inflict on the victim; and any special qualities (such as age or infirmity) of the victim himself. The presence or absence of a weapon also has significance, but mainly as a proxy for an assessment of the harm intended or the fear of the victim. Some states, though, ignore these factors to a great degree and punish many different degrees of assault similarly. Maryland, for example, allows all assaults to be punished by up to ten years' imprisonment, except for those resulting in serious physical injury or in which a weapon is used, which are punishable by up to twenty-five years' imprisonment. Vermont recognizes three assault offenses punishing acts that differ little with sentences that differ greatly: assault (causing "bodily injury"), punishable by up to one year's imprisonment; aggravated assault (causing "serious bodily injury"), punishable by up to fifteen year's imprisonment; and maiming ("cut[ting] or disabl[ing] a limb or member of another person"), punishable by seven years' to life imprisonment. Some states, oddly and redundantly, recognize these distinctions by enacting several offense provisions, but proceed to categorize the different provisions within the same grade. Some states are motivated by the three considerations above, but follow those considerations to recognize specific distinctions that are actually irrelevant. For example, several states consider as an aggravating factor whether the victim was the referee of a sporting event. Michigan establishes a penalty of up to life imprisonment for armed assault with intent to rob, and up to fifteen years' imprisonment for unarmed assault with intent to rob, but a maximum of ten years' imprisonment for assault with intent to maim, to "do great bodily harm less than murder," or to commit a felony other than murder or robbery. Only a paltry few states recognize appropriate differences based on relevant considerations and only relevant considerations. Texas, for example, grades assaults based on three factors: the culpability level of the offender; the level of resulting injury; and whether the victim was elderly, a child, disabled, a police officer, or a family member.

Four considerations seem germane in attempting to distinguish different degrees of the crime of arson: whether the arsonist has jeopardized anyone's safety in addition to damaging property; the type of property damaged; the value of the property damaged; and the extent of the damage to the property. Most of the states that use "value" as a factor in sentencing focus on the third listed factor (the original value of the property) rather than the fourth (the extent of the damage caused by the arson), which is perhaps more likely to be a suitable benchmark for gauging the seriousness of an offense. A handful of states recognize more than one of these considerations in making determinations of the proper punishment for an arson offense. Kansas, for example, provides for sentencing variations based on both the presence (and the extent) of a threat to human safety and on differences in the extent of the damage done to property. Many other states, however, fail to recognize these distinctions or impose improper distinctions. Maryland has a maximum punishment of five years' imprisonment for committing arson on anything other than a dwelling or other structure, but a punishment of up to ten years imprisonment for threatening to commit arson on a structure. The District of Columbia punishes arson committed on one's own property with intent either to defraud or injure more severely than any other type of arson. West Virginia, along with several other states, defines one punishment for arson regardless of whether the building burned was occupied, unoccupied, or even vacant, and barely distinguishes different types of buildings. Vermont overcompensates for its refusal to distinguish between occupied, unoccupied, or vacant buildings by imposing first-degree murder liability in relation to any arson "by means of which the life of a person is lost." Nevada's very definition of arson assures oversimplified grading of punishment: "[a]ny person shall be deemed to have 'set fire to' a building, structure or any property mentioned in [the arson provisions] whenever any part thereof or anything therein shall be scorched, charred or burned."

Gradations of punishment for theft generally look, quite reasonably, to the object of the theft. Grading distinctions may be based either on the stolen item's value or its type. With respect to the former, it should always be an improvement for a state to recognize more value categories, since this ensures that the punishment will be increasingly narrowly tailored to reflect the harm caused by the theft - at least, to the extent that value lost equals harm caused. Indeed, some states recognize a fairly large number of value categories. Texas, for example, grades punishment according to seven value levels ($200,000 or more; $100,000-$200,000; $20,000-$100,000; $1,500-$20,000; $500-$1,500; $50-$500; and under $50). Additionally, some specific things may merit special concern when they are the objects of theft; for example, many states punish more severely when a firearm is stolen, regardless of its value.

A number of states, on the other hand, scarcely bother to consider the significance of the value of the item stolen; several recognize as few as two value categories, and others recognize none. Perhaps worst of all, some states recognize only two value categories but provide for hugely divergent punishments based on whether the value of the stolen item is above or below the line separating the two categories. For example, Maryland provides for a maximum of eighteen months' incarceration for theft of an item whose value is less than $300. For items whose value exceeds $300, however, the maximum punishment increases to fifteen years of prison time. Montana, Michigan, Rhode Island, West Virginia, and Vermont are substantially similar.

As for consideration of the type of item stolen, on the other hand, more does not always mean better as far as the recognition of distinctions is concerned. A number of states carve out specific punishments seemingly at random for the theft of particular items. Mississippi's code serves as an excellent - meaning, of course, a terrible - example. After distinguishing between grand larceny (stealing an item worth $250 or more - punishable by up to five years' imprisonment, a $1,000 fine, or both) and petit larceny (stealing an item worth less than $250 - punishable by up to six months' imprisonment, a $1,000 fine, or both), Mississippi creates particular punishments for shearing wool from a dead sheep, stealing milk from a cow, and stealing crabs or crab pots. Stealing timber with a value of less than $25 is punishable by a fine of $200 to $500, 30-100 days' imprisonment, or both; if the value exceeds $25, the punishment is a fine of $200 to $500, one to five years' imprisonment, or both. Theft of livestock, regardless of its value, must be punished by one to five years' imprisonment and a fine of $1,500 to $10,000. Stealing another person's dog is punishable by "a fine of not more than five hundred dollars, or imprisonment in county jail for not more than six months, or both, or imprisonment in the penitentiary for not less than one year nor more than two years."

F. Overall Ratings

The codes' overall scores, the total of the scores from each of the five questions, are as follows:

The codes fall into five roughly defined groups. In the first group fall the top seven codes, which all score reasonably well on each question. (We note that Tennessee and Kentucky, though falling slightly further down the scoring list, also score well across the board.) Second comes a group of codes, Hawaii through Pennsylvania, that perform well on most of the questions but have unusually low scores on one or another question -- usually question 5, and sometimes question 4. Most of the states in a third group, between Wisconsin and Washington, are weak in two or more questions. Their scores often decline steadily as they move from question 1 to 5, or at least are significantly higher for questions 1 and 2 than for questions 3, 4, and 5. (Interestingly, the Pacific Northwest states of Oregon and Washington both defy this trend, and any other trends exhibited in our scoring). In a fourth group, from Ohio to North Carolina, the eleven codes perform poorly on most questions but have one or (occasionally) two higher scores, more often than not on question 5. Finally, the nine worst codes receive uniformly low scores on every question.

Upon examination, a general trend reveals itself. The vast majority of the codes that are neither consistently excellent nor consistently awful perform better with respect to the criteria that evaluate the rules of conduct than those that evaluate the rules of adjudication. Nearly all of the codes between Hawaii and Georgia on the list above fall within this trend. The best of these codes -- Hawaii, New Jersey, North Dakota, and possibly Delaware -- appear to replicate the failings of the Model Penal Code as well as its successes, as they receive much lower scores for question 5 than for the other four questions. Forming an exception to the general rule that codes score well on our conduct questions, but poorly on our adjudication questions is the relatively small group of states -- Ohio, Nebraska, Virginia, Florida, New Mexico, and North Carolina -- that score well only on question 5. These states appear to have taken seriously the need to punish fairly and consistently once a violation has been established, but have neglected the task of defining the violations themselves; they poorly define, or fail to define, the acts that are crimes or the actors that are criminals.

Yet we must point out two caveats to prevent the reader from translating the codes' scoring performance into a generalized impression that American criminal codes are good at conduct rules (questions 1 and 2), but bad at adjudication rules (questions 3, 4, and 5). First, the scoring criteria for our conduct rule questions tend to be more relativistic, and somewhat more forgiving of minor flaws, than are the corresponding adjudication rules criteria. For example, some codes may receive our maximum score for "drafting style" or "readability" because they are much easier to understand than other codes, but this does not mean that they lack any significant room for improvement in this regard. A truly objective and demanding benchmark of clarity and concision would reveal these codes to be far from perfect. Unfortunately, such a benchmark is difficult to devise, and anyway, since the current project is comparative in nature, it would gain little from uniformly reducing all of the codes' scores for a given factor.

Second, not all of the codes exhibit the tendency to perform well with respect to conduct rule questions and poorly with respect to adjudication rule questions. As one would expect in any analysis of the enactments of fifty-two different legislatures, some of the scoring results reveal idiosyncracies in some codes. New York, Oregon, South Dakota, and Nevada all receive much lower scores for question 1 than for any other question. Kansas has a high score for question 5 but a correspondingly low one for question 3. Connecticut, Oklahoma, and Idaho all have scores for question 3 that deviate from their usual performance. But surely the most unusual scoring pattern is that of Washington, the only state that performs significantly better on the questions related to adjudication rules than on the questions related to conduct rules.

The codes' scoring averages on the questions posed by our study suggest two observations. First, American criminal codes (excepting the handful of codes that seem to do everything poorly) tend to fare better in addressing the quantitative "comprehensiveness" issues addressed by our questions 1 and 3 (which essentially ask, "is everything there?") than the more qualitative concerns of questions 2, 4, and 5 (which essentially ask, "is everything good -- or at least, consistent?"). Scores for the codes in the top quartile and top half for questions 1 and 3 were at least 0.20 points higher than their counterparts for the three other questions. Second, in dealing with the adjudicative aspects of their codes, few states have devoted equal attention both to issues of establishing criminal liability and to issues of grading the punishment of those held liable.

These observations give rise to obvious prescriptions for state legislatures. The states whose codes are at the bottom of our rankings, when making decisions about reforming their code, should give priority to strengthening their performance on question 1 and 3 issues -- increasing comprehensiveness both in their statement of rules of conduct and, especially, their rules of adjudication. Comprehensiveness is both a fundamental necessity and -- because, as a general matter, it demands addition of new provisions rather than examination and amendment of old ones -- a more straightforward project than the projects suggested by other questions.

The states at the top, on the other hand, should focus on the concerns addressed by questions 2, 4, and 5. That is, they must strive to improve the clarity of their rules of conduct and to ensure that their already-existing adjudication rules are thorough, appropriate, and reflect the community's moral consensus about the blameworthiness of conduct. Having done this, those states should turn their attention to refining their system of grading offenses to better approximate the proper level of punishment.

Of course, even if such reforms were to occur, the improvements would be fleeting if later changes to the code were not tailored to the code's structure and style. If nothing else, our study reveals that the creation and maintenance of a sound code demands two things: expertise and vigilance. A team of criminal-law specialists was needed to draft the Model Penal Code, and even that code, in our view, is not flawless. As experience confirms, there is even less chance that state legislators will develop a genuinely good code. However sound their qualifications or strong their dedication may be, legislative drafters are rarely even aware of, much less worried about, the special needs of criminal codes. Even the best of codes will slowly deteriorate without active oversight, as later generations of legislators unfamiliar (or unconcerned) with the original code's scheme add, delete, and revise provisions, oblivious to the effects of these changes on the code as a whole. History indicates that the quality of the American criminal codes enacted in the wake of the Model Penal Code's promulgation has been eroded by subsequent modifications that disrupt the codes' initial clarity and coherence.

For these reasons, we advocate the institution of standing commissions to generate and monitor states' criminal codes. The idea, extreme though it may sound, is hardly without precedent. England has such a commission, as do many of the British Commonwealth countries. Most of the states that have adopted a variation of the Model Penal Code formed such a commission to draft the code, but dissolved the commission once the task was complete. We believe that these commissions could and should be kept alive. The cost of a standing commission would be low if its membership served only part-time and for low reimbursement (and perhaps were only compensated for their expenses). In any case, the advantages of a standing commission would far outweigh these costs, given the significance of the criminal code as both an instrument of the state's power and an expression of its values. The very immediacy and import of criminal law render it all the more susceptible to mere politicking rather than deliberative craftsmanship. A renewed and dedicated attention to criminal code reform may yet enable us to design codes whose quality is commensurate to the breadth and gravity of their social role.

 

State: ____________________________________ Total Score: _______

Scorer: __________________________________ Date: ________________

I. Communicating the Law's Commands

1. Does the code contain a comprehensive statement of the law's commands? ____ (0-4)

Scores: 4 = fully comprehensive (MPC rates 3.8); 3 = some holes, but generally comprehensive; 2 = many holes, but more codified than not; 1 = mix of codified provisions and case law references; 0 = depends almost entirely on case law or common law rules

Factors

A. Are all criminal offenses defined by the code? If not, how many are not? Does the code have a provision (like MPC § 1.05(1)) that bars the prosecution of common law offenses or other offenses not defined by a statute of the state?

AA. If prosecution of uncodified crimes is allowed, maximum allowed score is 2. More points off if uncodified offenses can be punished seriously (e.g., more than a year in prison).

B. Are the terms used in the code of conduct defined in the code? (also, but of less importance: Are the definitions in a place or one of a limited number of places such that one could reasonably know where to look for them?)

C. Are offenses defined incompletely, requiring reference to case law to fully determine the offense's requirements? If so, how often, and how serious are the offenses?

D. Are offenses defined by statutes outside the criminal code or criminal procedure code other than minor regulatory offenses? If so, how many and how serious? Some points added back if criminal code contains a cross-reference to these no-criminal-code offenses. [Less important than factors A-C]

E. Does the code (a) define the affirmative duties for which a person may be criminally liable for a failure to perform, or (b) does it incorporate by reference statutory duties defined by statutes outside the code, or (c) does it allow liability for duties imposed by law other than statute (as MPC § 2.01(3)(b) allows)?

EE. For (b), points off (very common); for (c), more points off and max of 3.8

F. Are all justification defenses codified in the code? (most states recognize justification defenses of self-defense, defense of others, defense of property, law enforcement, persons with special responsibility) Points off for justification defenses defined only in relation to a specific offense or group of offenses.

FF. If most justification defenses are undefined, max of 2.

Explanation/Comments:

2. Does the code effectively communicate the law's commands to the public? ____ (0-4)

Scores: 4 = effective communication; 3 = some weakness, but still good communication (MPC rates 3.2); 2 = many problems; 1 = poor communication; 0 = completely ineffective communication

Factors

A. Does the drafting style enhance the code's understandability by lay persons? plain words? short, clear sentences? section titles that accurately summarize the section?

B. Does the organization of the code enhance the lay person's understanding? do related sections appear near each other? do similar offenses appear next to each other? [does it contain a table of contents and an index? cross-references to related sections?]

BB. If random or alphabetical listing of offenses, max of 2.

BBB. If no distinct general part, max of 1.

C. Are the code's justification defense rules sufficiently simple that they reasonably can be remembered and applied in the factual situations in which they are likely to arise?

D. Does the code have overlapping offenses?

E. Within groups of related offenses, are offenses arranged in order of seriousness?

F. [How is the criminal code made available to the citizens of the jurisdiction? Does the state make any effort to educate its citizens as to the provisions of the criminal code and what they mean?]

Explanation/Comments:

[Q3 goes to legality interests in the principles of adjudication. The qualities of comprehensiveness and accessibility are desirable because they increase uniformity and predictability in application and decrease the potential for abuse. Q4 and Q5 go to blameworthiness interests.]

3. Does the code provide a comprehensive and accessible statement of its rules for determining whether to impose liability and, if so, the general grade of punishment to be imposed? ____ (0-4)

Scores: 4 = fully comprehensive (MPC rates 3.8); 3 = some holes, but generally comprehensive; 2 = many holes, but more codified than not; 1 = mix of codified provisions and case law rules; 0 = depends almost entirely on common law or case law rules

Factors

A. Does the code define the terms it uses in its adjudication provisions? culpability terms? does it limit the number of culpability terms that it uses in the definition of offenses? (does it make clear the hierarchical order of the culpability levels?)

B. Are all excuse defenses codified? (most states have disability excuses of immaturity/infancy, insanity, duress, and involuntary act, and mistake excuses of mistake as to a justification ("justified if believes ..."))

Are all nonexculpatory defenses codified? (most states recognize nonexculpatory defenses of statute of limitations and entrapment)

Points off for defenses defined only in relation to a specific offense or group of offenses.

BB. If most excuses and non-exculpatory defense are undefined, max of 2.

C. Are the code's adjudication rules sufficiently detailed to insure uniform application to similar cases? (e.g., as in detailed justification rules of MPC)

D. Are the basic adjudication provisions (other than general defenses) fully codified? (e.g., provisions governing complicity, causation, consent, mistake, voluntary intoxication, and limitations on multiple offenses.)

Explanation/Comments:

4. Does the code accurately assess who does and who does not deserve criminal liability? ____ (0-4)

Scores: 4 = code very accurately assesses criminal liability to those who deserve it, and exempts from liability those who do not; 3 = code generally accurate, but a few problems; 2 = code has some problems in accurately assessing liability (MPC rates 3.4); 1 = code has many serious problems in accurately assessing liability; 0 = code entirely unreliable in assessing liability

Factors

A. Are there gaps in the code's criminalization scheme such that code fails to criminalize conduct that most of the community would believe sufficiently condemnable to deserve criminal conviction (e.g., a general negligent homicide offense)?

B. Does the code criminalize trivial offenses? (e.g., cutting in line) Does it codify a general defense for a de minimis infraction (e.g., MPC § 2.12)? Does the code contain unenforced offenses (e.g., adultery, sodomy)?

C. Does the code set proper minimum requirements for criminal liability? (e.g., MPC's "substantial step" requirement for attempt may be too thin a conduct requirement)

D. Does the code provide a minimum culpability level that is to be "read in" when an offense definition is silent as to culpability (such as MPC § 2.02(3))? Does the code provide a presumption against interpreting a statute as one of strict liability, unless the legislative intent to impose it is clear? When strict liability is imposed, is the punishment limited to civil-like penalties, such as a fine?

DD. If code provides negligence as the baseline for offense culpability, then, max of 2.5.

E. Does the code recognize a full set of excuse defenses? (usually included are insanity, duress, infancy/immaturity, involuntary act; excuses sometimes omitted include: involuntary intoxication, mistake due to reliance upon official misstatement of law, and mistake due to unavailable law)

EE. Give extra points for a code that recognizes a general excuse defense for a reasonable mistake of law (e.g., New Jersey § 2C:2-4(c)(3)).

F. Does the code recognize a full set of justification defenses? (usually included are self-defense, defense of others, law enforcement authority; justifications sometimes omitted include: lesser evils/necessity defense)

G. How many unjust liability rules does the code contain, and how central are they in the code's operation? (e.g., guilty but mentally ill, unconstrained felony murder)

Explanation/Comments:

5. For those offenders held criminally liable, does the code accurately assess the proper grade of punishment the offender deserves? ____ (0-4)

Scores: 4 = code very accurately assesses general range (grade) of punishment deserved; 3 = code generally accurate, but a few problems in assessing proper grade (MPC rates 2.0); 2 = code has some problems in accurately assessing proper grade; 1 = code has many serious problems in accurately assessing proper grade; 0 = code entirely unreliable in assessing proper grade

Factors

A. Are different grades of an offense based on appropriate factors? Does the code give proper weight to the grading factors it recognizes?

AA. Is resulting harm taken into account in grading? Or, are inchoate offenses graded the same as completed offenses? If all inchoate offenses are graded the same as completed offenses, as in MPC § 5.05(1), max of 2.

B. Does the code use offense grading categories, rather than providing a specific sentence for each offense? If so, how many categories does it have.

BB. If code does not use grading categories, max of 2.5.

BBB. If code has at least eight total offense categories (for felonies, misdemeanor, and violations), then no points off. Add points for more categories.

C. Are the groups of offenses in the same offense category generally similar in seriousness?

Explanation/Comments:

Total score ______ (0-20)

STATE: ____________________________________ TOTAL SCORE: _______

SCORER: __________________________________ DATE: ________________

1. Does the code contain a comprehensive statement of the law's commands? ____ (0-4)

Scores: 4 = fully comprehensive (MPC rates 3.8); 3 = some holes, but generally comprehensive; 2 = many holes, but more codified than not; 1 = mix of codified provisions and case law references; 0 = depends almost entirely on case law or common law rules

Factors

A. Are all criminal offenses defined by the code? If not, how many are not? Does the code have a provision (like MPC § 1.05(1)) that bars the prosecution of common law offenses or other offenses not defined by a statute of the state?

[test offenses = murder, assault, theft/larceny/burglary/robbery offenses, sexual offenses, offenses against family, arson, fraud offenses, inchoate off, possession offenses, kidnapping offenses, influence offenses, falsification offenses, obstruction of. government offenses, abuse of office offenses]

6) code contains provision barring prosecution of common law offenses,

5) code contains no provision dealing w/ whether/not common law offenses allowed, and all of the offenses we expect to be in a code (i.e. test offenses) are there,

4) code contains no provision dealing w/ whether/not common law offenses allowed, and most of the offenses we expect to be in a code (i.e. test offenses) are there except some offenses we expect to be in a code are missing,

3) code contains a provision explicitly allowing prosecution of common law offenses, but the code also explicitly limits the punishment for common law crimes or stipulates that only minor level (e.g. misdemeanor level) common law offenses are still in effect,

2) code contains provision allowing prosecution of common law offenses & the code either allows such offenses to be punished severely, or is silent as to any limit on punishment of common law offenses,

1) most (50% plus) of the offenses we expect to be in a criminal code are not present in this one (regardless of whether or not the code contains any provision barring the prosecution of common law offenses)

AA. If prosecution of uncodified crimes is allowed, maximum allowed score is 2. More points off if uncodified offenses can be punished seriously (e.g., more than a year in prison).

B. Are the terms used in the code of conduct defined in the code? (also, but of less importance: Are the definitions in a place or one of a limited number of places such that one could reasonably know where to look for them?)

[check for general definition section; test offenses = homicide, assault, theft, arson, sexual offenses]

5) code has general definition section, as well as definition section for most offenses (e.g. offense v. person)

4) code has general definition section, as well as definition section for many offenses (e.g. offense v. person)

3) code contains no general definition section but there are many offense specific definition sections

2) code contains a general definition section, but either has no offense specific definition sections or paltry few

1) code contains no general definition section, and no definition section for particular offense sections (or only paltry few at best)

C. Are offenses defined incompletely, requiring reference to case law to fully determine the offense's requirements? If so, how often, and how serious are the offenses?

[test offenses = homicide, assault, theft, burglary, larceny, arson]

4) of our test offenses, most are defined completely by the code

3) of our test offenses 50% plus are defined completely

2) of our test offenses less than 50% are defined completely

1) of our test offenses all are defined incompletely

D. Are offenses are defined by statutes outside the criminal code or criminal procedure code other than minor regulatory offenses? If so, how many and how serious?

[Some points added back if criminal code contains a cross-reference to these no-criminal-code offenses. Less important than factors A-C]

3) never (or at least we did not find any) are there offenses defined by statutes outside the criminal code

2) only minor regulatory offenses are defined by statutes outside the criminal code

1) offenses that should be in the criminal code are defined/found in statutes outside the criminal code

E. Does the code (a) define the affirmative duties for which a person may be criminally liable for a failure to perform, or (b) does it incorporate by reference statutory duties defined by statutes outside the code, or (c) does it allow liability for duties imposed by law other than statute (as MPC § 2.01(3)(b) allows)?

4) the code defines all the affirmative duties for which person may be liable,

3) the code incorporates by reference duties defined by statutes outside the code (even if a few duties are defined by the code),

2) code makes no mention of affirmative duties (even if a few duties are defined by the code),

1) code allows liability for duties imposed by law other than statute (even if a few duties are defined by the code)

EE. For(1), max of 3.8

F. Are all justification defenses codified in the code? Points off for justification defenses defined only in relation to a specific offense or group of offenses.

[possible justification defenses include self-defense, defense of others, defense of property, law enforcement, public duty, persons with special responsibility, lesser evils/necessity]

4) all justification defenses we expect in a good criminal code are defined by this code,

3) more than 50%, of the justification defenses we expect in a good criminal code are defined by this code,

2) less than 50% of the justification defenses we expect in a good criminal code are defined by this code

1) the code contains no justification defenses, or justification defenses are defined only in relation to specific offenses

FF. If most justification defenses are undefined, max of 2.

Explanation/Comments:

2. Does the code effectively communicate the law's commands to the public? ____ (0-4)

Scores: 4 = effective communication; 3 = some weakness, but still good communication (MPC rates 3.2); 2 = many problems; 1 = poor communication; 0 = completely ineffective communication

Factors

A. Does the drafting style enhance the code's understandability by lay persons? plain words? short, clear sentences? section titles that accurately summarize the section?

[test offenses = murder, assault, arson, and theft]

5) universally contains no (or almost no) readability problems -- for our test offenses, the code uses plain words, short/clear sentences, and section titles accurately summarizing the section,

4) code sometimes has minor readability problems here and there - around 50% or so of our test offenses in this code contain plain words but lengthy/unclear sentences, or complex words and short/clear sentences, but the other 50% of the test offenses use plain words/short sentences

3) universally contains minor readability problems-- for almost all test offenses, the code uses plain words but lengthy/unclear sentences, or complex words and short/clear sentences,

2) code sometimes has major readability problems here and there -- 50% plus of our test offenses contain lengthy/complex words/phrases, and lengthy/complex/unclear sentences. The rest of the test offenses are either OK or are subject to minor readability problems

1) code universally contains major readability problems -- for almost all test offenses, code uses lengthy/complex words/phrases, lengthy/complex/unclear sentences, and misleading section titles

B. Does the organization of the code enhance the lay person's understanding? do related sections appear near each other? do similar offenses appear next to each other? [does it contain a table of contents and an index? cross-references to related sections?]

6) code contains related sections that appear next to one another, similar offenses that appear next to each other, and there is a general part

5) code contains a general part and either related sections appear next to one another and similar offenses do not, or similar offenses appear next to one another and related sections do not

4) the code contains a random or alphabetical listing of offenses but it does have a general part

3) the code contains related sections that appear next to one another, similar offenses that appear next to each other, BUT, there is NO general part

2) code contains NO (or paltry) general part and either related sections appear next to one another and similar offenses do not, or similar offenses appear next to one another and related sections do not

1) the code contains a random or alphabetical listing of offenses and has NO general part

BB. If random or alphabetical listing of offenses, max of 2.

BBB. If no distinct general part, max of 1.

C. Are the code's justification defense rules sufficiently simple that they reasonably can be remembered and applied in the factual situations in which they are likely to arise?

5) justification rules are clear and self contained - the codes justification rules use plain words and short/clear sentences, and never require reference to outside provisions

4) rules either have minor readability problems or minor self containment problems, but not both - the codes justification rules use plain words and short/clear sentences, and occasionally require reference to outside provisions <or> justification rules use plain words and lengthy/complex sentences (or complex words and short sentences), and never require reference to outside provisions, <or> justification rules use plain words and lengthy/complex sentences (or complex words and short sentences), and occasionally require reference to outside provisions

3) rules have serious readability problems but no self containment problems --- the codes justification rules use complex words and lengthy sentences, and never require reference to outside provisions <or> no real readability problems, but rules have serious self containment problems - i.e., the codes justification rules use plain words and short/clear sentences, and almost always require reference to outside provisions

2) justification rules have serious readability problems and self containment problems - the codes justification rules use plain words and lengthy/complex sentences (or complex words and short sentences), and almost always require reference to outside provisions, <or> justification rules use complex words and lengthy sentences, and almost always require reference to outside provisions <or> justification rules use complex words and lengthy sentences, and occasionally require reference to outside provisions

1) the code has no justification rules

D. Does the code have overlapping offenses?

[test offenses = murder, assault, theft, sexual assault]

4) no overlapping offenses found for our test offenses

3) some overlapping offenses found for our test offenses

2) a lot of overlapping offenses were found amongst our test offenses

1) each of our test offense categories contained many overlapping offenses

E. Within groups of related offenses, are offenses arranged in order of seriousness?

[test offenses = murder, assault, theft, sexual assault]

4) for our test offenses, all offenses within a grouping of offenses were arranged in order of seriousness (from most serious to least serious)

3) for our test offenses, most offenses within a grouping of offenses were arranged in order of seriousness (from most serious to least serious)

2) for our test offenses, many offenses within a grouping of offenses were not arranged in order of seriousness (or were arranged from least serious to most serious)

1) for our test offenses, many offenses within a grouping of offenses are not arranged in order of seriousness

Explanation/Comments:

[Q3 goes to legality interests in the principles of adjudication. The qualities of comprehensiveness and accessibility are desirable because they increase uniformity and predictability in application and decrease the potential for abuse. Q4 and Q5 go to blameworthiness interests.]

3. Does the code provide a comprehensive and accessible statement of its rules for determining whether to impose liability and, if so, the general grade of punishment to be imposed? ____ (0-4)

Scores: 4 = fully comprehensive (MPC rates 3.8); 3 = some holes, but generally comprehensive; 2 = many holes, but more codified than not; 1 = mix of codified provisions and case law rules; 0 = depends almost entirely on common law or case law rules

Factors

A. Does the code define the terms it uses in its adjudication provisions? culpability terms? does it limit the number of culpability terms that it uses in the definition of offenses? (does it make clear the hierarchical order of the culpability levels?)

4) code contains at least 4 (and hopefully not more than 6 or 7) defined culpability terms, with hierarchy noted (e.g., purpose, knowing, reckless, negligence)

3) code contains many (5+) culpability terms with no hierarchy noted

2) code contains only few (1 to 4) defined culpability terms, with no hierarchy noted

1) code contains no defined culpability terms

B. Are all excuse defenses codified? (most states have disability excuses of immaturity/infancy, insanity, duress, and involuntary act, and mistake excuses of mistake as to a justification ("justified if believes ..."))

Are all nonexculpatory defenses codified? (most states recognize nonexculpatory defenses of statute of limitations and entrapment)

Points off for defenses defined only in relation to a specific offense or group of offenses.

[check for statute of limitation, voluntary act/omission, intoxication, duress, entrapment, mistaken justification, insanity, immaturity, mistake of law excuse]

4) all excuse defenses we expect to find in a good criminal code are included/defined by this code

3)more than 50% of the excuse defenses we expect to find in a good criminal code are included/defined by this code

2) less than 50% of the excuse defenses we expect to find in a good criminal code are included/defined by this code

1) the code contains no excuse defenses

BB. If most excuses and non-exculpatory defense are uncodified, max of 2.

C. Are the code's adjudication rules sufficiently detailed to insure uniform application to similar cases? (e.g., as in detailed justification rules of MPC)

[check multiple offense limitations, causation, mistake, complicity, intoxication, consent, mental illness negating element, self-defense, defense of others, defense of property, law enforcement, public duty, persons with special responsibility, lesser evils/necessity, voluntary act/omission, intoxication, duress, entrapment, mistaken justification, insanity, immaturity, mistake of law excuse, statute of limitation]

4) for our test list of adjudication rules, all are sufficiently detailed to insure uniform application - i.e. there are no short blurbs likely requiring case law development

3) for our test list of adjudication rules, more than 50% are sufficiently detailed to insure uniform application - i.e. there are very few short blurbs likely requiring case law development

2) for our test list of adjudication rules, less than 50% are sufficiently detailed to insure uniform application - i.e. they are all short blurbs likely requiring case law development, or the code simply does not contain those provisions

1) for our test list of adjudication rules, none are sufficiently detailed to insure uniform application - i.e. they are all short blurbs likely requiring case law development, or the code simply does not contain those provisions (i.e. there are no justification/excuse defenses, and culpability terms are ad hoc/made up on an offense by offense basis)

D. Does the code contain a full set of the provisions (other than excuses) needed for adjudication of violations?

[check for multiple offense limitation, causation, mistake, complicity, intoxication, consent, mental illness negating element]

4) code contains all of the general adjudicatory provisions we expect to find in a good criminal code

3) code contains more than 50% of the general adjudicatory provisions we expect to find in a good criminal code

2) code contains less than 50% of the general adjudicatory provisions we expect to find in a good criminal code

1) code contains none of the general adjudicatory provisions we expect to find in a good criminal code

E. Does the code contain a section that sets general rules for how its provisions are to be interpreted?

2) code contains rules by which it is to be interpreted

1) code does not contain rules by which it is to be interpreted

Explanation/Comments:

4. Does the code accurately assess who does and who does not deserve criminal liability? ____ (0-4)

Scores: 4 = code very accurately assesses criminal liability to those who deserve it, and exempts from liability those who do not; 3 = code generally accurate, but a few problems; 2 = code has some problems in accurately assessing liability (MPC rates 3.4); 1 = code has many serious problems in accurately assessing liability; 0 = code entirely unreliable in assessing liability

Factors

A. Are there gaps in the code's criminalization scheme such that code fails to criminalize conduct that most of the community would believe sufficiently condemnable to deserve criminal conviction (e.g., a general negligent homicide offense)?

[test offenses = murder, assault, theft/larceny/burglary/robbery offenses, sexual offenses, offenses against family, arson, fraud offenses, inchoate off, possession offenses, kidnapping offenses, influence offenses, falsification offenses, obstruction of. government offenses, abuse of office offenses]

3) no gaps in code's criminalization scheme noted

2) a few gaps in code's criminalization scheme noted

1) many gaps in code's criminalization scheme noted

B. Does the code criminalize trivial offenses? (e.g., cutting in line) Does it codify a general defense for a de minimis infraction (e.g., MPC § 2.12)? Does the code contain unenforced offenses (e.g., adultery, sodomy)?

[check for de minimis; test offenses = indecency offenses, sex offenses]

5) code contains no trivial/unenforced offenses and a de minimis defense is included (although not having one shouldn't be a problem here - note also that this basket is an ideal that likely no code meets)

4) code contains few trivial/unenforced offenses and a de minimis defense

3) code contains few trivial/unenforced offenses and no de minimis defense

2) code contains many trivial/unenforced offenses and a de minimis defense

1) code contains many trivial/unenforced offenses and no de minimis defense

C. Does the code set proper minimum requirements for criminal liability? (e.g., MPC's "substantial step" requirement for attempt may be too thin a conduct requirement)

[check inchoate offenses]

3) code provides for attempt/conspiracy/inchoate offenses definition that makes clear that mere preparation is insufficient for the actus reus element

2) code provides no general definition of attempt/conspiracy/inchoate offenses

1) code explicitly provides for substantial step (or some equivalent) for attempt/conspiracy/inchoate offenses

D. Does the code provide a minimum culpability level that is to be "read in" when an offense definition is silent as to culpability (such as MPC § 2.02(3))? Does the code provide a presumption against interpreting a statute as one of strict liability, unless the legislative intent to impose it is clear? When strict liability is imposed, is the punishment limited to civil-like penalties, such as a fine?

[check culpability strict liability and provisions]

4) the read-in provision contained in the code is recklessness or knowledge

3) no read-in provided is contained in the code

2) the read-in provision contained in the code is explicitly negligence

1) the read-in provision contained in the code is explicitly strict liability

DD. If code allows strict liability to be assumed in absence of stated culpability term, max of 2.5.

E. Does the code recognize a full set of excuse defenses?

[usually included are insanity, duress, infancy/immaturity, involuntary act; excuses sometimes omitted include: involuntary intoxication, mistake due to reliance upon official misstatement of law, and mistake due to unavailable law]

4) all excuse defenses we expect to find in a good criminal code are included/defined by this code

3)more than 50% of the excuse defenses we expect to find in a good criminal code are included/defined by this code

2) less than 50% of the excuse defenses we expect to find in a good criminal code are included/defined by this code

1) the code contains no excuse defenses

EE. Give extra points for a code that recognizes a general excuse defense for a reasonable mistake of law (e.g., New Jersey § 2C:2-4(c)(3)).

F. Does the code recognize a full set of justification defenses?

[usually included are self-defense, defense of others, law enforcement authority, defense of property; justifications sometimes omitted include lesser evils/necessity defense, public duty, special responsibility, military orders]

4) all justification defenses we expect in a good criminal code are defined by this code,

3) more than 50%, of the justification defenses we expect in a good criminal code are defined by this code,

2) less than 50% of the justification defenses we expect in a good criminal code are defined by this code

1) the code contains no justification defenses, or justification defenses are defined only in relation to specific offenses

G. How many unjust liability rules does the code contain, and how central are they in the code's operation? (e.g., guilty but mentally ill, unconstrained felony murder)

5) code contains no felony murder rule, and no guilty but mentally ill provision

4) code contains constrained felony murder rule, and no guilty but mentally ill provision

3) code contains unconstrained felony murder rule, and no guilty but mentally ill provision

2) code contains constrained felony murder rule, BUT has a guilty but mentally ill provision

1) code contains unconstrained felony murder rule, AND has a guilty but mentally ill provision

Explanation/Comments:

5. For those offenders held criminally liable, does the code accurately assess the proper grade of punishment the offender deserves? ____ (0-4)

Scores: 4 = code very accurately assesses general range (grade) of punishment deserved; 3 = code generally accurate, but a few problems in assessing proper grade (MPC rates 2.0); 2 = code has some problems in accurately assessing proper grade; 1 = code has many serious problems in accurately assessing proper grade; 0 = code entirely unreliable in assessing proper grade

Factors

A. Are different grades of an offense based on appropriate factors? Does the code give proper weight to the grading factors it recognizes?

[test offenses = assault, arson, and theft. Considerations applied to test offense of assault are: the harm the assailant inflicted on the victim, the harm the assailant intended to inflict on the victim, and any special qualities (such as age or infirmity) of the victim himself. Considerations for arson are: whether the arsonist has jeopardized anyone's safety in addition to damaging property, type of property damaged, value of the property damaged, and value of the damage to the property. Considerations for theft are the stolen item's value or its type.]

5) generally recognizes appropriate factors

4) generally recognizes appropriate factors; sometimes applies factors crudely or uses irrelevant ones

3) frequently applies factors crudely

2) frequently applies factors crudely; sometimes ignores relevant factors or uses irrelevant ones

1) frequently ignores relevant factors or uses irrelevant ones

AA. Is resulting harm taken into account in grading? Or, are inchoate offenses graded the same as completed offenses?

[check inchoate offense, particularly attempt; caveat: conspiracy often is graded same as substantive offense because group criminality is seen as an independent harm]

6) specific provision: inchoates lower than complete

5) specific provision: inchoates lower than complete, but with exceptions

4) no specific provision; apparently few or no inchoates treated same as complete

3) no specific provision; some or many inchoates treated same

2) specific provision: inchoates = complete, but with exceptions

1) specific provision: inchoates = complete

If all inchoate offenses are graded the same as completed offenses, as in MPC § 5.05(1), max of 2.

B. Does the code use offense grading categories, rather than providing a specific sentence for each offense? If so, how many categories does it have.

[check for classes of offenses provision]

4) code contains more than 10 grading categories

3) code contains 8-10 grading categories

2) code contains less than 8 grading categories

1) code contains no grading categories - uses specific sentences for offenses

BB. If code does not use grading categories, max of 2.

BBB. If code has at least eight total offense categories (for felonies, misdemeanor, and violations), then no points off. Add points for more categories.

C. Are the groups of offenses in the same offense category generally similar in seriousness?

4) For all grades searched & all the offenses listed, groups of offenses in the same grading category were generally similar in seriousness

3) under emphasizing seriousness - lower grade (like C, D, E felony, or misdemeanor) categories contained some offenses sufficiently serious to be in a higher grading category

2) overemphasizing seriousness - higher grading categories contained some offenses that were not that serious

1) for all grades searched the offenses listed were not generally similar in seriousness - failings include both over emphasizing and under emphasizing seriousness -- higher grading categories contained some offenses that were not that serious, and lower grade (like C, D, E felony, or misdemeanor) categories contained some offenses sufficiently serious to be in a higher grading category

Explanation/Comments:

Total score ______ (0-20)

Appendix C: Summary of Scoring Documentation, By State