The Nature of Criminal Law Conventions

The unwritten, binding, and evolving norms that supplement statutory criminal law.


In this post, I want to explore two features of criminal law conventions in more depth: custom and obligation.

There are two broad theories of common law. The predominant theory today, influenced by the legal realists, is that common law is a form of judge-made law—that is, judges promulgate common law through their rulings. Someone has to make the law. As Justice Holmes famously claimed, "The common law is not a brooding omnipresence in the sky but the articulate voice of some sovereign or quasi-sovereign that can be identified …."

But there is a much older tradition (about which Steve Sachs has written a great article) asserting that judges find the law rather than promulgate it. This tradition views the common law as consisting of binding customs that society recognizes, even though no authority has promulgated it. These customs form through a decentralized process. And as Sachs explains in his article, legal customs are not unique; we routinely recognize "familiar norms of fashion, etiquette, or natural language" as binding even though no one has authoritatively pronounced them.

Criminal law conventions resemble customary law in this latter sense. Like the common law, the customs are defined despite the lack of formal promulgation by a person acting with legislative or judicial authority, they evolve over time, and they are widely understood by the public and by those in authority as binding.

To illustrate these features, I give three categories of examples in the article: desuetudinal laws, traffic offenses, and the boundaries of federal criminal law. Start with desuetudinal crimes, which are crimes that have long been widely and openly ignored (e.g., adultery). The existence of these crimes presents a challenge to our concepts of legality. On the one hand, they are statutory law, and in most jurisdictions, no internal legal rule renders these laws unenforceable. On the other, there is widespread recognition that the conduct prohibited by these laws is not unlawful de facto. When former New York Gov. David A. Paterson was confronted with evidence of adultery, he declared publicly, "I didn't break the law"—which he could say honestly, even though adultery theoretically remains a misdemeanor in New York.

A second example consists of traffic offenses. Virtually all people who drive commit traffic offenses—probably every time they are on the road. Speed limits are set too low. The general recommendation is that speed limits should be set to the 85th percentile speed of traffic. Real-world speed limits are about 10 to 15 miles per hour underneath that. Ultimately, executive nonenforcement compensates. For example, in Virginia courts in 2018, only 40 of 590,000 speeding cases involved drivers speeding less than five miles per hour over the limit. (It is unclear whether these tickets are actually for de minimis speeding or if some officers reduced the ticket.) We all know that the real speed limit is about 10 to 15 miles per hour over the real limit, and police largely respect the convention.

A third example concerns the boundaries of state and federal criminal law. As many scholars have criticized, the number of federal crimes has exploded over the past 140 years, and it now heavily duplicates state criminal law. But real federal criminal law is not that pervasive. Contemporary federal criminal law heavily involves immigration, drugs, guns, and fraud. Federal statutes against domestic violence, carjacking, guns in school zones, and motor vehicle theft largely go unenforced.

Within these offenses, there are widely understood norms dividing federal and state jurisdiction. The federal government generally prosecutes drug trafficking, but leaves drug possession to the states. The federal government also prosecutes cases in which prosecutors believe more severe federal sentences are warranted. In felon in possession gun cases, for example, federal prosecutors generally target those who have more significant criminal histories or who may be a danger to the community.

Much like common law, these customs evolve as society's norms evolve, and in many cases, these norms developed when someone violated the old norms. Until the 1970s, gun crimes were treated as state offenses, even though federal law began regulating ordinary firearms in 1938. In the 1980s and '90s, federal prosecutors more aggressively prosecuted federal gun crimes, overcoming resistance from federal judges that their courtrooms were being "transformed into a minor-grade police court." Gun cases are now a large fraction of the federal criminal docket, and that trend has been relatively stable.

We can see that evolution in progress today. The federal government, for example, generally does not interfere with the possession of marijuana for medical use when allowed by state law, even though possession remains (for now) a federal crime. Sanctuary cities, whether for immigration or guns, are an effort to establish broader decriminalization norms. Statutory law can be difficult to amend or repeal. But conventional criminal law, much like common law, continually evolves.

The other critical feature of criminal law conventions is that they are accepted as binding in some form. While it may be easy to write off much of what I have written above as just reflecting prosecutorial discretion, I think that kind of rationalization fails to capture how deeply some norms are held and the consequences for violating them.

Take the Gun Free School Zones Act example that I gave in my previous post. Just like one could ask whether the Queen can veto a bill, one could ask whether an assistant United States Attorney could prosecute an off-duty police officer for unlawfully carrying a firearm in a school zone. As with the British constitutional example, the formal legal answer is yes. If the assistant U.S. Attorney brought such charges, the judge could not dismiss it because he thought the law was bad or stupidly applied in an individual case. Yet, the real-world answer is likely no. If a U.S. Attorney filed such a charge, the political blowback from the law enforcement community would be so intense that any such prosecution would be dropped and the politically appointed U.S. Attorney could very well be fired. Even if the case somehow got to trial, there would be a substantial risk that the jury would acquit notwithstanding the evidence. Given this reality, federal prosecutors do not try to bring such charges regardless of their personal views of the merits.

Or take the speeding example. One might claim that in a world where everyone speeds, the police have to direct their resources towards the most serious violators. But the executive branch does not pursue all technical speeders even where no such resource constraints exist. The District of Columbia, for example, has widely deployed speed cameras. Just like they now ticket every red-light violator at places where red-light cameras are installed, the District could automatically ticket every driver who exceeds the speed limit by one mile per hour or more. But the District does not; as a matter of course, speed camera tickets are not issued for speeding within 10 miles per hour of the limit. The District respects the conventions on traffic from a sense of obligation and with recognition that some form of non-legal sanction may result if they are not followed.

There is, however, one critical difference between criminal law conventions and constitutional conventions of the kind existing in the British system. Constitutional conventions are rigid rules for which deviation would be a high-publicity event. If the Queen vetoed one bill, no matter how unimportant, the entire country would notice and it would provoke a constitutional crisis.

The same is not true for criminal law conventions. Deviations from conventions are often low-information events that target unsympathetic individuals.  If police began strict enforcement of speed limits, they (and the community's elected officials) may suffer political consequences. But nothing stops police from engaging in isolated targeting of drivers. In many cases, these may be pretextual stops to search for evidence of other crimes.

I will have more to say about the pretextual issue in my last post when I discuss how we should shape legal doctrine around a primarily conventional criminal law system. For now, I will end by flagging that, unlike constitutional conventions, criminal law conventions are susceptible to isolated, low-cost deviations, which trigger fairness and rule of law concerns.