Sign In
    Wisconsin Lawyer
    March 04, 2010

    Editorial: Defining "Voluntary Confession"

    Which police interrogation tactics render suspects’ confessions involuntary and therefore inadmissible? The line between a voluntary and an involuntary confession is not as clear as one might assume, because the legal standard courts use to draw the line is too dull. The author proposes a revised test.

    Richard J. Sankovitz

    Wisconsin LawyerWisconsin Lawyer
    Vol. 83, No. 3, March 2010

    Since 2007, when it became Wisconsin policy for police officers to record the interrogations of all felony suspects,1 the state’s circuit courts have been getting a closer and more frequent look at what goes on inside police interrogation rooms. The recordings give courts a play-by-play look at the sometimes-subtle, sometimes-intimidating, sometimes-clever tactics wielded by skillful interrogators to try to persuade suspects to confess. Defendants listening to their recorded confessions often shake their heads in disbelief. Many sincerely believe that they must have been tricked.

    The recording of confessions definitely has yielded at least one of the benefits it promised: courts are presented with many fewer motions to suppress. On the other hand, the motions that are filed are more challenging because courts have not previously examined these interrogation tactics at such close range.

    What makes the motions even more challenging is that courts are not presently equipped with a legal standard that reliably sorts interrogation tactics that are acceptable from those that are not. The concern in every interrogation is whether the interrogation has yielded an involuntary confession. Involuntary confessions are not only inadmissible; too often they are false. In hindsight, courts may be left questioning whether a suspect has confessed for “the wrong reason.” The law should guide courts to spot an involuntary confession and exclude it.

    The line the law currently draws between voluntary and involuntary confessions, however, is too dull to reliably identify involuntary confessions. I believe Wisconsin courts need a legal test with a sharper edge. My belief is based on my experience as a judge presiding in a felony court routinely deciding motions to suppress confessions and in particular on having listened to more than 40 recorded confessions over the past 18 months.

    Current law. Under current law, a confession is considered voluntary, and therefore admissible, if it is the product of a “free and unconstrained will, reflecting deliberateness of choice.”2 Wisconsin courts have been applying this standard for more than 40 years, without any real refinement.3

    A sharper test. Rather than try to look inside a suspect’s head or heart to judge whatever subjective impression a particular interrogation tactic may have made, I propose that courts should look objectively at the interrogation tactic itself. The court should judge whether the interrogator employed a tactic that foreseeably caused the suspect to consider any reason for confessing other than the suspect’s actual guilt or innocence or the lawful consequences of confessing or not confessing. The test assumes that the police apply some pressure to the suspect to confess, and that such pressure yields a confession, but the focus of the test is not on the amount of the pressure but rather its subject matter. This would make for a surer test of whether a suspect has confessed for “the wrong reason.”

    In the discussion that follows, I spell out a revised standard and demonstrate with examples how it would apply in practice.

    You Make the Call

    Consider three hypothetical confessions. In each example, the suspect is an adult of normal intelligence and composure. (Therefore, a separate branch of the law of voluntariness having to do with suspects who are unusually vulnerable to police pressure, as a result of youth or of physical or mental disability for example, does not come into play here.4)

    In each example, consider what happens just before the suspect confesses:

    Example 1. The interrogator informs the suspect, a parolee with a history of committing violent crimes who has been to prison twice, that if he confesses, his case will stay in state court and the district attorney and the judge will consider his cooperation, which could lead to a shorter sentence. But if he doesn’t, the U.S. attorney will take over the case and, if the federal judge follows the sentencing guidelines, the suspect will be facing at least 10 years in prison.

    Example 2. The interrogator informs the suspect that “we got the whole thing on video from a dash cam in an unmarked squad car,” which is a complete fabrication.

    Example 3. The interrogator tells the suspect she has only a limited time to come clean because detectives in another interview room are talking to the suspect’s companion, who may already be confessing. The interrogator states that whoever confesses first will get the best deal from the prosecutor.

    In each example, the suspect confesses and is charged with a crime and then moves to suppress the confession. The defendant argues that the confession was coerced and therefore must be excluded from evidence at trial.

    How would you rule? Was the confession the product of a “free and unconstrained will, reflecting deliberateness of choice”? Did any of the suspects have a real choice, given the stark options or misinformation or threats supplied by the police interrogator? Does it seem likely in these examples that an average suspect’s will to resist giving a false confession will be overborne?

    Shortcomings of the Current Voluntariness Standard

    Judging voluntariness even in the simplest of cases can be tricky. Although the law requires confessions to be voluntary, the law does not require that they be volunteered. Perhaps it is more accurate to say that courts currently require merely that confessions “not be involuntary.” But we are still left scratching our heads; if courts cannot map what “voluntary” means, then courts cannot be certain when a confession has crossed the border into “involuntary.”

    Examples like those above make one think about the utility of the legal standard courts currently employ. The critical terms – deliberateness of choice and free and unconstrained will – are too dull, in my opinion, to distinguish voluntary from involuntary confessions, for the following reasons:

    The first main flaw. The first main flaw in the current test is that the deliberateness-of-choice component is substantially under-inclusive (that is, under this part of the test very few confessions can be excluded), and the free-and-unconstrained-will component is equally over-inclusive (under this part of the test, too many confessions must be excluded).

    Hon. Richard J.   Sankovitz

    The Hon. Richard J. Sankovitz, Harvard 1983, is a Milwaukee County Circuit Court judge.  

    Consider the deliberateness-of-choice component. Just about every confession, false or not, can be said to be the result of a deliberate choice. The following might seem to be an obvious situation for excluding a confession: a suspect who chooses to confess rather than be beaten. But does the deliberateness-of-choice test dictate its exclusion? The suspect is given a choice. The choice is voluntary, at least in the sense that the suspect makes the choice and that the choice is not made for the suspect. And, the choice is deliberate, at least in the sense that the suspect thinks it over before choosing. Thus, under the terms of this test the confession arguably is admissible.

    The suspect, however, has not been given a real choice – but nowhere does the standard mention illusoriness as a factor to be considered, and in no case in Wisconsin has this factor been analyzed in such terms. Everyone can agree that a choice made under threat of physical violence is involuntary. Nevertheless, the legal standard courts have been given to sort out voluntary from involuntary confessions does not dictate this result.

    Under the other component of the test – whether the suspect’s confession was the product of a free and unconstrained will – hardly any confession should be admitted, at least if the words of the standard are taken at their face value. Use of the words “free and unconstrained” could be interpreted to mean that the suspect’s will to resist confessing cannot be constrained at all. Thus, the standard would seem to bar any tactic that works to overcome a suspect’s resistance or that results in a change of mind or a change of heart.

    Now it must be said that the rule has never been applied so literally. Judges do not routinely throw out confessions simply because the police have been successful in getting a suspect to change his or her mind. Judges recognize not only that suspects often change their minds (if only to avail themselves of the benefits of confessing), but also that suspects’ minds do not simply change on their own. Judges recognize that suspects often confess because police interrogators are called in to get them to confess. Police interrogators are specifically trained to create a favorable climate for confessing and are practiced in bending a suspect’s will.

    The U.S. Court of Appeals for the Seventh Circuit expressed this same view in United States v. Rutledge:

    “Taken seriously [the standard we employ to judge the voluntariness of confessions] would require the exclusion of virtually all fruits of custodial interrogation, since few choices to confess can be thought truly ‘free’ when made by a person who is incarcerated and is being questioned by armed officers without the presence of counsel or anyone else to give him moral support. The formula is not taken seriously…. In any event, very few incriminating statements, custodial or otherwise, are held to be involuntary, though few are the product of a choice that the interrogators left completely free.”5

    Perhaps because the free-and-unconstrained-will element is so inflexible on its face, over the years courts have felt the need to incorporate into it a balancing test. In almost the same breath as courts say the suspect’s will must be unconstrained, courts now also say that his or her confession will be admitted at trial if the “pressures brought to bear” on the suspect do not “exceed[] the [suspect’s] ability to resist.”6 But this amendment to the test does not make up for its shortcomings. In the case of an adult of normal intelligence and composure who confesses, the acid test of the person’s “ability to resist” is whether the person in fact succumbs to police pressure and confesses. Once the person has confessed, what tool is there to measure the reservoir of “ability to resist” that was not in fact applied?

    The second main flaw. The second main flaw in the current voluntariness standard is that it is premised on an obsolete notion of the human psyche. Courts consider the suspect’s will and speak in terms of whether his or her will has been “overborne.”7 But, as the Rutledge court put it, “[t]here is … no ‘faculty of will’ inside our heads that has two states, on and off, such that through careful reconstruction of events the observer can determine whether the switch was on when the defendant was confessing.”8

    A Practical Test of Voluntary Confession

    When the Seventh Circuit critiqued the voluntariness standard in Rutledge, it sketched out an alternative. In Rutledge and its progeny, the Seventh Circuit has asked whether there is evidence of “extrinsic considerations that would overcome [the suspect’s] will by distorting an otherwise ration-al choice of whether to confess or remain silent.”9 As long as the suspect, in choosing whether to confess, is not forced to “consider anything beyond his own beliefs regarding his actual guilt or innocence, his moral sense of right and wrong, and his judgment regarding the … evidence linking him to the crime,”10 the suspect’s confession should not be deemed involuntary.

    This test was used in one published appellate opinion in Wisconsin. In State v. Triggs,11 a police officer told a suspect, untruthfully, that eyewitnesses saw her shoot at her ex-boyfriend. The court of appeals held that the misrepresentation was innocuous because it “did nothing more than corroborate [other] truthful information” the officer discussed with the suspect.12

    Wisconsin should adopt a similar standard to replace the ineffectual “free and unconstrained will, reflecting deliberateness of choice” standard. The new standard might be framed like this:

    “If the police pressure a suspect to confess, and if confession was the result of the police pressure, the confession is deemed involuntary if the foreseeable effect of the police pressure was to cause the suspect to consider some factor ‘extrinsic’ – that is, not germane – to the suspect’s actual guilt or innocence or the lawful consequences of confessing, not confessing, thinking it over more, gathering more information, or otherwise remaining silent.”

    Putting the New Standard to the Test

    This more refined standard would equip circuit courts to home in on the salient features of confessions such as those hypothesized above.

    Consider the common tactic of threatening federal prosecution, as illustrated in the first example. The tactic is not condemnable merely because it involves a threat. Every interrogation involves some degree of threat. At the very least there is the threat, often made explicitly, that if the suspect refuses to confess, he or she will lose the benefits of confessing. As long as there are benefits to confessing, suspects will always feel compelled to preserve them and will feel threatened if they seem in jeopardy.

    A threat of federal prosecution might be particularly condemnable, though, if the threat caused a suspect to give a false confession for reasons having nothing to do with his or her guilt or with the lawful consequences of not confessing – for example, if the U.S. attorney only rarely prosecutes crimes like those alleged against the suspect. On the other hand, if crimes like those alleged against the suspect are regularly prosecuted, and if the suspect’s potential exposure is as the interrogator describes it, then the interrogator simply presented the suspect with a legitimate choice.

    Consider deceitfulness, as was used in the second example. In my experience listening to recorded confessions, it is not uncommon for interrogators to engage in deception when rooting out a confession. Most often, the deception is limited and consists only of exaggerating the strength of the evidence against the suspect. In other cases, it is more extreme. In about a quarter of the confessions I reviewed, the interrogator suggested the presence of evidence – witness accounts, photographs, fingerprints, and so on – that did not exist.

    It is generally held, however, that deception by itself does not render a confession involuntary.13 This is because the suspect knows whether he or she did what he or she is accused of doing and is in an unassailable position to judge the truthfulness of the interrogator’s claims, or at the very least to demand that the interrogator show him or her the evidence. As long as interrogators focus their fiction on subjects germane to the suspect’s actual guilt or innocence, a confession should not be deemed coerced. Contrast such a situation with Lynumn v. Illinois, the archetypal interrogation deceit case, in which the suspect was led to believe she would lose her children and her welfare benefits if she did not cooperate.14

    Finally, consider whether a confession should be excluded because the interrogator pits one suspect against the other, as in the third example. A suspect who felt compelled to confess in this circumstance would have a legitimate complaint because it is not lawful for an interrogator to promise leniency. Such a promise would not be germane to the suspect’s actual guilt or innocence or the lawful consequences of confessing or not. Under the revised standard proposed in this article, this confession would not be admissible at trial.


    To meet the challenges of motions to suppress confessions in the era of recorded interrogations, courts need a sharper tool for sorting out voluntary from involuntary confessions. Focusing less on the amount of pressure applied, and instead on whether the interrogator employed factors extrinsic to the suspect’s actual guilt, or on the lawful consequences of confessing or not confessing, would better sort out confessions made for the right reason from those made for a wrong reason. When presented with a case in which the voluntariness of a confession is at issue, the Wisconsin Supreme Court should consider adopting this revised test.


    1Wis. Stat. § 968.073.

    2See, e.g., State v. Hoppe, 2003 WI 43, ¶ 36, 261 Wis. 2d 294, 661 N.W.2d 404.

    3The standard was suggested by the U.S. Supreme Court in Haynes v. Washington, 373 U.S. 503, 514 (1963), and adopted in Wisconsin in State v. Hoyt, 21 Wis. 2d 284, 294, 128 N.W.2d 645 (1964).

    4E.g., Hoppe, 2003 WI 43, 261 Wis. 2d 294.

    5900 F.2d 1127, 1129 (7th Cir. 1990).

    6Hoppe, 2003 WI 43, ¶ 36, 261 Wis. 2d 294.

    7See, e.g., State v. Knapp, 2003 WI 121, ¶ 98, 265 Wis. 2d 278, 666 N.W.2d 881, vacated on other grounds, 542 U.S. 952 (2004).

    8Rutledge, 900 F.2d at 1128.

    9Holland v. McGinnis, 963 F.2d 1044, 1051 (7th Cir. 1992). See also Rutledge, 900 F.2d at 1129.

    10Holland, 963 F.2d at 1051.

    112003 WI App 91, 264 Wis. 2d 861, 663 N.W.2d 396.

    12Id. ¶ 20.

    13See Frazier v. Cupp, 394 U.S. 731, 739 (1969) (“The fact that the police misrepresented the statements that [the co-defendant] had made is, while relevant, insufficient … to make this otherwise voluntary confession inadmissible”); United States v. Montgomery, 555 F.3d 623, 632 (7th Cir. 2009); Holland, 963 F.2d at 1051 (false claim that witness had seen defendant’s vehicle at crime scene); Lucero v. Kerby, 133 F.3d 1299, 1311 (10th Cir. 1998) (misrepresentation regarding fingerprint evidence found in victim’s home); Ledbetter v. Edwards, 35 F.3d 1062, 1066, 1070 (6th Cir. 1994) (falsified fingerprint-comparison chart, claim that three witnesses had identified defendant, and staged victim identification); Triggs, 2003 WI App 91, ¶ 15, 264 Wis. 2d 861.

    14Lynumn v. Illinois, 372 U.S. 528 (1963).      

Join the conversation! Log in to comment.

News & Pubs Search

Format: MM/DD/YYYY