In the Supreme Court term that ended last month, the Court decided United States v. Davis, applying the void-for-vagueness doctrine. The decision in Davis reaffirms an unremarkable proposition. If one cannot readily determine whether and when a criminal law applies to people’s behavior, then the law is unconstitutional. In Davis, the Court said that one cannot tell from 18 U.S.C. § 924(c) whether a given crime qualifies as a “crime of violence”—which could make a defendant eligible for a harsher sentence. The phrase “crime of violence” is therefore unconstitutionally vague. People must be able to know from the text of a criminal statute what it prohibits and what it threatens.
This column asks when the void-for-vagueness doctrine makes sense. The due process component of vagueness doctrine—a right to notice—implicitly rests on a theory of human behavior. When that theory accurately describes what people do, then the doctrine can promote justice. When theory and behavior diverge, however, the doctrine may yield unwarranted leniency and solve a nonexistent problem.
A Theory of Human Behavior
The apparent theory behind the due process (fair notice) objection to vague criminal statutes is that people are price-sensitive. If they want a product and we raise its price, their demand for the product will drop. If we lower the price, in turn, people will more readily consume the same product. This is the law of supply and demand in economics.
How, you might ask, does that bear on the criminal law? Think of a crime as a product that the criminal wants to consume. Say you very much enjoy breaking into people’s houses and stealing their things. You like the excitement of the activity, and you enjoy the rewards of always finding something different that you can just take without having to pay for it. If there were no law prohibiting burglary and larceny, then we could expect you to regularly indulge in this exciting break from the ordinary.
Assume that a legislature passes laws prohibiting burglary and larceny precisely because people like you would otherwise burglarize and steal. It imposes a penalty on these crimes to attach a price to the otherwise-attractive activity. Once you hear that breaking into homes and stealing people’s things are crimes that carry jail or prison sentences, you begin to think twice about committing them. The threat of prison acts as a price for the consumer goods of burglary and larceny. When we say that the criminal law deters and thereby reduces crime, we mean exactly this—punishing behavior puts a price on it. Absent a price, demand will predictably rise.
So far so good. We can reasonably anticipate an increase in crime if the crime carries no consequences. Furthermore, in order for a would-be offender to respond appropriately to the incentives, he must know that his actions will result in a penalty. A secret criminal law will have no deterrent effect. As far as someone contemplating crime is concerned, an unknown law is little better than no law at all.
What about fairness? Does theorizing a market in criminal conduct tell us anything about due process? Yes, it does. If we believe that people might respond to threatened penalties by conforming their conduct to the law, then we have an obligation to let people know what we are threatening to do if they transgress. This is essential to being able to fairly punish them later for proceeding with their criminal plans. Those who do not realize that we have put a price on burglary and larceny will be far more likely to commit these acts and can reasonably say that they did not have enough information to make the right choice. Understood differently, we effectively induce people to commit crimes when we lead them to believe that no punishment will follow. We generally should not punish people for failing to act on secret information.
From a fairness point of view, the most unjust situation would involve punishing someone for a crime that does not exist in the law. Say police decided to arrest someone for looking at the sun during a solar eclipse, as President Trump did two summers ago. Assume that the criminal law says nothing to suggest that this conduct is against the law (as opposed to incredibly stupid). To lock up the star gazer would be unjust because we did nothing to warn him that his actions would have consequences (beyond the risk of blindness). Warnings are essential to fairness. This principle extends well beyond the criminal law to the tort area, where pharmaceutical companies must warn consumers of side effects and risks and where informed consent is part of a medical professional’s duty to the patients that she treats.
Vague criminal laws act much like secret or nonexistent criminal laws in that they do not effectively tell people what they must avoid doing. A prohibition against “conversation about trifling topics on Mondays” (set aside the First Amendment infirmity) will leave good people perplexed and nervous about what this law demands of them.
Where things become less clear is in the area of penalties. At the extremes, of course, people attend to penalties and make rational consumer choices on the basis of them. If you learned that leaving your home in the next two months was a crime punishable by execution at the guillotine, you would probably stay home. And if you learned instead that the same crime went with a fine of fifteen cents, you would continue to go outside without fear.
Once we are talking about years in prison, though, it seems unlikely that the difference between twenty and twenty-one is going to make a difference. Are there people who would rob a convenience store if they thought the punishment was twenty years but think better of the whole scheme upon learning the penalty was a year longer? There is reason to think that at some point, differences in penalties might make no difference in behavior. Does anyone imagine that criminals make it a point to find out what the punishment is for stabbing someone in the street or setting fire to a neighbor’s house? Is there a proper price point that will deter the knife-wielding attacker?
Anything is possible, and we lack a clean natural experiment. When crime rates rise and fall, numerous confounding variables typically frustrate attempts to make confident attributions. I suspect, however, that the relevant question for many aspiring criminals is not
“how many years will I get?” but “how likely am I to be caught?” Someone who felt certain that the police and prosecutors would find him and try him would refrain from committing the offense, even if the penalty was “only” two years in prison. This may be why an increased police presence, according to many, helped bring down the crime rate in New York City—an officer at your back virtually guarantees detection.
Once detection is uncertain, I would predict that the criminally inclined become optimistic about their odds of getting away with their offenses. They believe they will not get caught and therefore may feel relieved of having to focus too heavily on the penalty for those who are apprehended. If I am correct, then increasing penalties will boost the prison population without actually motivating people to obey the law.
Beyond implementing a misguided social policy, in such a case, we would also be conferring an unwarranted benefit on criminals if were to hold some laws “void for vagueness” for failure to give notice. If people contemplating crime do not respond to sentence length differences—perhaps once we go beyond ten, fifteen, or twenty years—then the fact that a criminal law was unclear or vague about the conditions under which an additional five or ten years would follow the regular sentence might not truly be unfair to a defendant. If I am indifferent, in other words, to whether my crime might result in fifteen, twenty, or twenty-five years behind bars, then a lack of clarity in informing me that the penalty really is twenty-five rather than fifteen does not actually affect me or my decision whether to commit the crime. By hypothesis, I would not have acted any differently had I realized (through clearer drafting) that the penalty was as long as it was.
Do I think that the void for vagueness doctrine should therefore shrink to account for such cases? Probably not. It is difficult to tell ex ante whether people will care about the variation between possible sentences for their crimes. And if there is even one person who would have cared and would have refrained from committing the crime if the threatened penalty were clearer, then it is unfair to her, at least, to punish her as though she had knowingly incurred the cost of her actions. Even if only she develops a reliance interest on the penalty clause of the law, due process belongs to individuals, not just populations. What we can do, though, if I am correct about incentives, is cut back quite a bit on the length of sentences. If, as I propose, sentence length does not affect people’s behavior very much, then long sentences as an attempt to deter crime are a waste of resources. As I have explained, it is not that most criminals deserve better. It is that we do. We are the ones paying a price in this scenario for a product—deterrence—that we aren’t getting. The rational thing is to stop paying.
