Reaching Equilibrium: How Do Criminal Law Conventions Develop?


Justice Scalia's celebrated dissent in Morrison v. Olson, which argued that the independent counsel statute was unconstitutional, contains two important themes. The first concerned separation of powers and explained why Congress had unconstitutionally removed an officer exercising purely executive power from the supervision of the President. But Part V of the opinion heads in a different direction: checks and balances. Quoting Justice Jackson, Justice Scalia explained, "With the law books filled with a great assortment of crimes, a prosecutor stands a fair chance of finding at least a technical violation of some act on the part of almost anyone." To curtail prosecutors' expansive formal legal power, Justice Scalia understood that "the primary check against prosecutorial abuse is a political one."

While statutory criminal laws may be broad, they are narrowed by unwritten criminal law conventions. And criminal law conventions develop because executive officials face political sanctions for violating them.

Observe, for example, the dispute over federal enforcement of marijuana laws. Between 2009 and 2013, the Department of Justice instructed U.S. Attorneys not to prosecute those growing and using marijuana lawfully under state law. At first, the guidance was applied to medical marijuana, but the Department later expanded it to cover recreational marijuana in states that had legalized it. Five years later, President Trump was in office. His Attorney General, Jeff Sessions, opposed marijuana legalization and rescinded the memo. The political backlash was swift. Senator Cory Gardner—a Republican from Colorado, which had legalized recreational marijuana—responded by blocking the confirmation of about twenty Department of Justice nominees. With a member of his own party refusing to confirm nominees, President Trump promised (despite his duty to "faithfully execute the laws") that the Department of Justice would not interfere with Colorado's marijuana industry. Thus, even though Sen. Gardner could not successfully get Congress to amend or repeal the federal law on marijuana, he attained de facto decriminalization by raising the political price of enforcement beyond what the executive branch was willing to pay.

Prosecutors face significant pressures from a multitude of directions. Legislatures may pass broad and vague criminal laws, which empower prosecutors. But then legislatures curb that power by controlling enforcement agencies and threatening legal reform if prosecutors become overzealous. Legislatures also strip power from prosecutors who underenforce the law. Larry Krasner, the Philadelphia District Attorney, has come under severe criticism for his proposed leniency in cases involving violent crimes and gun crimes. The Pennsylvania legislature responded by giving the state Attorney General concurrent jurisdiction to prosecute certain gun crimes in Philadelphia, thereby weakening Krasner's ability to set criminal justice policy in the city.

Many prosecutors face electoral pressures. In some jurisdictions, this makes it difficult for prosecutors to be excessively lenient. More recently, however, excessive enforcement has become politically toxic, with many cities electing "progressive prosecutors." Federal prosecutors are appointed, not elected. But their conduct affects the president's political fortunes, and many U.S. Attorneys harbor ambitions for higher political office. So they are often responsive to voters' concerns.

Juries make it difficult for prosecutors to bring cases against sympathetic defendants or defendants who, although they may have technically breached some law, have not violated community norms. Faced with improperly brought cases, jurors may nullify. And even if they do not nullify, jurors may review the evidence in the light most favorable to a sympathetic defendant. Although few cases today go to juries, enforcement decisions take place in the shadow of the jury. The threat that the defendant may take the case before the jury influences and disciplines how prosecutors act. (See, for example, Anna Offit's article Prosecuting in the Shadow of the Jury.)

Judges have a variety of tools to combat prosecutorial abuse. Some are legal. Faced with vague or overbroad laws, judges may construe them to contain strict mens rea requirements, by narrowing the actus reus, or by striking them down as unconstitutional. A few jurisdictions have adopted Section 2.12 of the Model Penal Code, which gives judges power to dismiss "de minimis infractions." Appellate judges may also be more apt to reverse convictions of sympathetic defendants for trial errors that, for less sympathetic defendants, may be written off as harmless. And judges have soft-power tools. They can encourage prosecutors to drop charges or plead them out at reduced sentences. In many cases, a judge may lack the legal power to stop a determined prosecutor from seeking a maximum sentence. But prosecutors are repeat players who are incentivized not to anger judges before whom they appear.

Finally, the expansion of federal criminal law paradoxically may curb some prosecutorial abuses. Duplicative federal criminalization can make it difficult for local officials to be too lenient. Krasner's leniency led the Philadelphia U.S. Attorney to indict some violent-crime defendants under harsher federal law. Excessive harshness or leniency can also trigger "uncooperative federalism" among federal and state agencies. Federal and state law enforcement are interdependent. State governments have far more police and prosecutors, while the federal government has sophisticated investigative resources and provides grants to state and local governments. Prosecutors who defy norms may strain this relationship.

Given the variety of checks on prosecutorial power, one question that emerges with unwritten conventions is whether the conventions are national or local in nature. I think there are some of both. Unwritten English law recognized both "general customs," which Blackstone called "the common law, properly so called" and "particular customs" which Blackstone described as unwritten laws that "affect[ed] only the inhabitants of particular districts." In our federal system, I think criminal law conventions have a similar structure. Some conventions are national—e.g., the decriminalization of consensual sex offenses. Others may be local, such as cities that widely tolerate some personal use of recreational drugs or refuse to enforce certain gun crimes.

Having multiple layers of customs promotes development in the law. As local governments experiment with new norms, some local customs (e.g., drug decriminalization) may spread nationally. In other cases, national norms can override pernicious local customs. For example, in the 1960s, the federal government, faced with local legal systems in the South that widely tolerated violence against African Americans, began prosecuting some racially motivated violence under federal civil rights laws. Duplicative federal-state criminalization, thus, provided a countermeasure against local customs that left African Americans vulnerable to violence by underenforcing the law.

In summary, indirect sanctions discipline prosecutors and curb the effective scope of their discretion. To restate a point that I made in my last post, I do not think that current criminal law conventions solve all that ails our criminal justice system. Obviously, it does not. My final post tomorrow will tackle some of those issues. But the existence of considerable constraints on prosecutorial discretion does much to mitigate the potential harms of our imperfect statutory criminal law.

NEXT: Gains and Losses: The Arbitrary Effect of Decision Frame

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. “But the existence of considerable constraints on prosecutorial discretion does much to mitigate the potential harms of our imperfect statutory criminal law.”

    We’re all in agreement that this is a good thing, right?

    1. Yes, although the obvious question becomes, how does one formalize this so that prosecutors can’t be selectively constrained in a racist or otherwise biased manner. The answer I would propose is well, statutory law, but these posts apparently disagree? But idk what is being proposed.

    2. Not always.

      For instance, I suspect in the Jim Crow South, these norms were deployed to make prosecutions of extrajudicial white violence against Black people (e.g., lynchings) impossible.

      Another example is the OJ Simpson case. Probably the biggest reason he got away with double murder was because the DA, Gil Garcetti, felt beholden to and listened to people who wanted him to get away with double murder. I did a tweetstorm on this:

      1. Dilan Esper….That was a tweetcane, not tweetstorm. You stopped at 100. The tweetcane tells quite a story. What motivated you to tell the story on August 23rd of 2020?

      2. The problem with the theory — besides the complete lack of anything resembling actual evidence for it — is that it involves Garcetti deliberately tanking the prosecution to secure black votes by acting in ways that voters were unlikely to realize involved deliberately tanking the prosecution. It assumes that he thought that a black voter would reward him for seemingly vigorously prosecuting OJ but failing.

        Also, claiming that the motive for killing Goldman was to eliminate a witness is made up.

        (Also, of course, the claim that it would be “unconstitutional” in the federal system to try black people in the suburbs isn’t correct; the 6th Amendment applies to judicial districts, not individual divisions.)

  2. I remember an editorial against the initiative petition. The voters will occasionally pass some popular measure against the will of the legislature, and with that outlet available become less likely to throw out politicians who do not represent their interests.

  3. Baloney. All of your restraints exist almost entirely in theory. For the average matter, that is, 99% of the cases handled by a prosecutor, there are no restraints.

    1. Pretty much. An entire class of people, in a visible and repeated manner, have to be screwed over by a prosecutor before anyone even notices. Or one really, really big mistake, but even then, the consequences are very minimal.

  4. Blackstone’s distinction between local and national customs (which, if it’s the matter that I’m remembering, is phrased in terms of liberties rather than norms) makes an important qualitative distinction as well. Local liberties always have a provenance, dating to a certain grant or a recorded custom, and they’re zealously guarded and remembered. National liberties (which prevail in the US) are a more amorphous thingamajig. Both, though, are questions of law, rather than the equitable discretion that this article seems to be describing.

    Mr. D.

Please to post comments