The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
For the last four years, I have been serving as an Adviser to an ongoing American Law Institute project, Principles of the Law: Policing. This week, the ALI membership is considering key parts of the Principles at its Annual Meeting in Washington, D.C. I can't make it to the Annual Meeting, so I figured I would explain here some of my concerns with the Principles. My concerns leave me lukewarm on the project, including some of the key parts to be considered this week.
Let me start with the big picture. The ALI is perhaps best known for its Restatements of the law and the Model Penal Code. The Principles is more like the Model Penal Code than Restatements. It is an effort to develop normative principles of how police agencies should do their jobs. That's a tricky goal because the law of policing in the United States is primarily federal constitutional law—and in particular Supreme Court interpretations of the Fourth and Fifth Amendments. Federal constitutional law sets the floor, and legislatures, state courts, and policing departments are then free to add additional restrictions as they feel appropriate.
The basic goal of the Principles is to set out ways in which the ALI thinks that the police should restrict themselves beyond what existing law requires. And I think it's fair to say that the Principles reflects the views of its Reporters, and selected Advisers, that the Supreme Court's interpretations of the Constitution don't go nearly far enough in restricting police practices. The Principles therefore set out to recommend a set of best practices that they feel the police should follow that go beyond the restrictions that the Supreme Court has required under the Constitution.
That basic idea of recommending non-constitutional limits on policing practices is certainly commendable. Indeed, the role of non-constitutional limits on policing is a theme of my own academic work. But my main difficulty with the Principles is that it tries to impose a one-size-fits-all answer on some really hard problems. It reflects a particular view of what limits would be appropriate for all policing, and it recommends that all police departments adopt those limits.
My own view of policing is more cautious. I'm not sure that one-size-fits-all works beyond the constitutional floor. There is an incredibly broad range of police agencies in the United States. There are federal agencies, state agencies, and local agencies. Some agencies are enormously large, and others are very very small. Some have broad mandates, and others have very narrow ones. In some areas, like eyewitness identification, we have social science that leads to useful rules to recommend. But in most areas, we don't. And I'm skeptical that it works to recommend a single set of rules for every agency outside that context.
My concerns often leave me unsure of whether a significant number of the rules proposed by the Principles are wise. Some of the rules strike me as sound. Others are probably useful in some settings. A few of the rules may be good guidance in some settings but I think go too far in other settings.
All of this leaves me playing Justice Harlan to the Reporters' criminal procedure revolution, unable to be on board some of the key recommendations that the Principles make.
This is all rather abstract, so let me turn to examples. The draft under consideration this week is 162 pages long, and the ALI may threaten legal action if I post it, so unfortunately I can't show the full draft. But I want to run through the provisions that most implicate the concerns I have. I'll start with some of the sections on search and seizure. I'll then turn to the Miranda interrogation rules.
Section 4, The Search and Seizure Rules
Let's start with Section 4.03, which discusses when the police should stop and arrest individuals and how they should interact with people when they do. This section reflects the concerns of procedural justice theory, which focuses on perceptions of police legitimacy and trust in policed communities. In light of those concerns, this section contends that police citizen encounters should generally be limited. "[A]gencies should limit the overall use of initial encounters, stops, and arrests, to circumstances in which they directly promote public safety and minimize harm to the public." And when officers take such actions, they should explain to individuals what they are doing and why to ensure that the police action is likely to be deemed fair:
Officers should establish the legitimacy of their encounters with members of the public by treating individuals with dignity and respect, explaining the basis for the officers' actions, giving individuals an opportunity to speak and be heard, and engaging in behaviors that convey neutrality, fairness, and trustworthy motives.
This is usually good advice. But should it always be required? Procedural justice is important, but I'm not sure it needs to apply to every police citizen interaction to achieve its benefits.
Similarly, Section 4.04 limits police questioning during a Terry stop to inquiries "necessary to investigate the crimes or violations for which the officer has reasonable suspicion, or as necessary to ensure officer safety." The officer shouldn't stop a person for one reason and then ask questions that aren't necessary to investigate the matter that provided the constitutional justification for the stop.
This is done to prevent pretextual stops, the comments explain. Without such limits on questioning, "those secondary intrusions may themselves become the goal of the stop, leading to unnecessary and perhaps unnecessarily intrusive encounters between officers and the public." (The Reporter's Notes suggest that section is merely a summary of existing Fourth Amendment law, but that's not correct: The Supreme Court expressly rejected this rule in Muelher v. Mena. This limit on questioning would be a major shift in practice.)
I understand the concerns with pretextual stops. But I would think there are at least some circumstances when we would want them, and when we would want questioning outside the justification for the stop. For example, say there has been a string of recent robberies in the neighborhood, and the officer pulls over a car for speeding. The driver seems to match the rough description of the participants in the robbery, but there isn't enough evidence that creates proper reasonable suspicion to detain the driver for those crimes. Is it clear that we don't want the officer to ask any non-speeding questions in that situation?
Next, the black letter rule of Section 4.06 suggests that police agencies should consider adopting a rule that officers cannot ask for consent unless they first have reasonable suspicion. The comments to the rule replace this suggestion with a more categorical rule: "Officers should not seek consent to conduct a search unless they have reasonable suspicion to believe that the search will turn up evidence of a crime and unless they can explain to the individual why they would like to conduct a search." Again, this is often a good approach. But won't there be many circumstances when an officer has suspicions that don't amount to "reasonable suspicion"? And should we want officers to be categorically barred from asking for consent in those circumstances?
Similarly, Section 4.07, on search incident to arrest, would prohibit full searches of a person incident to arrest unless "there is reasonable suspicion to believe that the arrestee is concealing a weapon or evidence that would not be uncovered through a pat-down search." Again, the goal seems to be to avoid pretext searches: Agencies should enact policies on searches incident to arrest so that such searches "are not used as pretext to look for evidence of a crime or violation that is unrelated to the offense for which the individual was arrested." I agree that, in some cases, such pretexts will be harmful. But are they always harmful? I wouldn't think so.
Chapter 11, the Miranda/Interrogation Rules
The materials on interrogations take some steps that I find troublesome. Let me focus on two.
Section 11.03 discusses when Miranda warnings should be given. The rule states: "Officers should inform suspects of their right to refrain from answering and their right to counsel, and ensure that any waivers of those rights are meaningfully made." At first that seems unobjectionable, as it's basically Miranda law but with a higher pre-Berghuis waiver standard.
But if I'm reading the materials correctly, there appears to be something very different going on. As I understand Chapter 11, officers are supposed to inform people of these rights whether or not they are in custody, and indeed whether or not they are suspects at all. That's a rejection of the basic idea of Miranda, which was that warnings were required only when a suspect is in custody.
Here's what the comments say, in Section 11.01, about why, as the Reporter's notes put it, the Principles aim to "move beyond the unwieldy concept of custody" in interrogation law:
A detailed body of constitutional law applies to police questioning of suspects. One important area of constitutional law—the Miranda doctrine—draws a line by asking whether a person is deemed to be in "custody." See Miranda v. Arizona, 384 U.S. 436, 444 (1966). . . . The focus in constitutional law on the issue of "custody" can be quite formalistic, and remote from the concerns that motivate these Principles. For example, an innocent person who is not formally in custody still may face great pressure to confess falsely. A vulnerable person, such as a juvenile or mentally ill person, may receive unfair treatment that implicates concerns of legitimacy, even if not considered a suspect and not formally deemed to be in custody during the questioning. That said, the concerns with accuracy, coercion, and legitimacy may well be greater in the settings in which more formal custodial questioning occurs. No matter in what form or setting questioning occurs, police professionals ought to have an abiding interest in getting it right. Thus, these Principles do not take as their starting place the line between custodial and noncustodial interviews. Rather, the focus is on obtaining accurate statements with minimal coercion.
So if I'm reading the Principles correctly, everyone gets Miranda warnings. It seems that suspects and non-suspects, those in custody and those out of it, all receive the same warnings. And people questioned are required to be informed that they have a right to counsel to be present for questioning even when they have no such rights—as they don't unless they're being interrogated in custody, when Miranda provides it. Do we really want the police to always give Miranda warnings any time they ask questions to people? That seems a quite remarkable legal rule.
One last example. Section 11.04 explains how interrogations should be conducted. It states:
When questioning individuals, officers should:
(a) minimize the length of questioning;
(b) avoid leading questions and disclosing details that are not publicly known;
(c) avoid threats of harm to the individual or others or, conversely, avoid making promises of benefits to the individual or others;
(d) avoid the use of deceptive techniques that are likely to confuse or pressure suspects in ways that might undermine accuracy of evidence;
(e) ensure the individual has access to basic physical and personal needs, including food, water, rest, and restrooms; and
(f) not question the individual in an environment that is unduly uncomfortable.
The comments to the rule explains that the point is to minimize the coercive nature of police questioning, both to avoid false confessions and to respect the dignity of those questioned. And some of these are certainly good rules.
But others seem overly restrictive, at least assuming the requirement to "avoid" those facts is meant as a serious limit and not just a small thumb on the scale. For example, is it necessarily problematic for the police to ask leading questions? ("You killed him, didn't you?") Isn't it often useful and unproblematic to provide details that are not publicly known? ("Your husband was shot today in his driveway by someone at point blank range. I regret to tell you that he died. Do you know anything about it?") Are deceptive techniques that might "pressure" suspects necessarily problematic? ("Your friend just confessed and says you pulled the trigger, what do you have to say about that?")
Of course, interrogators should be aware of the risks that these techniques may pressure suspects in ways that lead to problematic interrogations. But those same techniques may be very effective, and I'm skeptical that the right line between permitted and banned is a general rule that those techniques should be avoided.
I don't know if my concerns will resonate with the broader ALI membership. But I thought I would raise them for members and others who may be interested.