Back in 1973, Harvard Law Professor Alan Dershowitz published an article, “Preventive Confinement: A Suggested Framework for Constitutional Analysis,” 51 Texas Law Review 1277. Forty years later, the article remains very relevant—so much so that Dershowitz built upon it in his 2006 book Preemption: A Knife that Cuts Both Ways. (I worked as a research assistant for Dershowitz for a few years during and after college, including on Preemption.)
In the article, Dershowitz described what he calls a “legal labeling game.” Here’s how the game is played: A legislature enacts a law that attempts to abridge somebody’s liberty. (Sex offenders are a salient example that I’ll address in detail in later posts.) To facilitate the process of imposing that liberty abridgment, the legislature makes few procedural safeguards available to the person who will suffer the loss of liberty. That person then goes to the courts and attempts to prevent the liberty abridgment by claiming that he is being punished, or being subjected to a criminal prosecution, without all the constitutional protections to which he is entitled. The state, in response, claims that the liberty abridgment is not part of punishment or criminal prosecution, but rather a civil, non-punitive, or regulatory action. Dershowitz wrote that the courts consistently accepted the state’s position and therefore helped them to win the legal labeling game.
As I expect will become clear, I’m sympathetic to this account, and I believe it continues to play out over and over today.
A leading case in determining whether a measure is civil or criminal is Kennedy v. Mendoza-Martinez, 372 U.S. 144 (1963). (The majority opinion was by Justice Arthur Goldberg, for whom Dershowitz clerked the following term, not coincidentally.) As we will see, deportation, removal, and exclusion are current hot topics at the intersection of civil law and criminal law. Thus, it’s appropriate that the key case from half a century ago also falls into that category.
The case involved a U.S. citizen, Francisco Mendoza-Martinez, who left the county to avoid serving in the military during World War II. Years after he returned, he was deemed to have forfeited his U.S. citizenship, and was ordered deported. The determination and order were automatic, without any trial or proceeding. The Court considered whether Mendoza-Martinez had been punished by deprivation of citizenship without due process. This, of course, hinged on whether the stripping of citizenship constituted “punishment”—in other words, the very question at the center of the legal labeling game Dershowitz identified. The majority opinion in Mendoza-Martinez lists “the tests traditionally applied to determine whether an Act of Congress is penal or regulatory in character”:
- Does the sanction stemming from that Act of Congress involve an affirmative disability or restraint?
- Has the sanction historically been regarded as a punishment?
- Does the sanction come into play only on a finding of scienter?
- Will the sanction’s operation promote the traditional aims of punishment—retribution and deterrence?
- Is the behavior to which the sanction applies already a crime?
- Can the sanction be assigned to an alternative purpose to which it may rationally be connected?
- Does the sanction appear excessive in relation to the alternative purpose assigned?
(Though I draw on the Court’s language in this list, I’ve omitted quotation marks to make it more readable.)
In this case, after considering the factors and “the objective manifestations of congressional purpose” for the statute in question, the Court found that stripping of citizenship was punitive and ruled in Mendoza-Martinez’s favor—an atypical example of the defendant coming out ahead in the legal labeling game. Unfortunately, the Court acknowledges that the factors it listed “may often point in differing directions,” and as later cases have made abundantly clear, the Mendoza-Martinez factors have not succeeded in providing a clear partition between the civil and criminal law.
In my opinion, the legal labeling game is alive and well, and later courts’ application of the Mendoza-Martinez factors shows that we continue to lack a well-defined distinction between civil law and criminal law.