Suspect Interviews and False Confessions

Interviewing in the United States

Leo (2008) points out that, contrary to popular myths of the American justice system, “the goal of police interrogation is not necessarily to determine the truth” (p. 23). Leo discusses the police interviewers' focus on obtaining a confession from suspects rather than on truth or fact finding. This focus on confessions has a long history, and obtaining a confession still remains an important goal of the psychologically oriented Reid technique (named after one of the original authors) that replaced the “third degree” (i.e., the use of physical force during interrogation, deprivation of food and sleep) in the United States. The Reid technique is the most popular and frequently used police interview technique in the United States (Inbau, Reid, Buckley, & Jayne, 2001). It encourages interviewers to use a two-stage process. The first stage is a nonaccusatory interview (known as a behavioral analysis interview, or BAI), in which general background information about the suspect is obtained, rapport and trust are built, and a determination is made about whether or not the suspect is lying about the offense. If the suspect is judged to be lying, then the interview progresses to a nine-step accusatory approach, typically referred to as “interrogation” (Inbau et al., 2001). In this article, we use the term interview to encompass both of the Reid technique stages. Masip, Herrero, Garrido, and Barba (2010) have identified fundamental problems with the BAI (i.e., lack of grounding in science, unfounded commonsense beliefs, and inaccurate recommendations).

Kassin and Gudjonsson (2004) argue that the nine-step Reid technique can be reduced to three general phases: “custody and isolation” (i.e., the suspect is detained and isolated, anxiety and uncertainty are generated in order to weaken resistance), “confrontation” (i.e., the suspect’s guilt is assumed and he or she is confronted with alleged incriminating evidence that may or may not be genuine; denials are rejected, even if they happen to be true, and the consequence of continued denial is emphasized), and “minimization” (i.e., the interrogator tries to gain the suspect’s trust and provides face-saving excuses for the crime, including suggesting that it was an accident or that the victim deserved it).

It is generally accepted that the Reid technique is effective in breaking down denial and resistance and eliciting confessions. The authors of the technique claim an 80% confession rate and state that, of the remaining 20%, “a small percentage of them could have been innocent” (i.e., they presume that a very high percentage of suspects they interrogate are guilty of the suspected offense; Inbau et al., 2001, p. 364). This claim of success has not been scientifically verified, and serious concerns about the technique’s association with false confessions have been raised (Kassin, 1997). In view of this, some authors (Bull & Soukara, 2010; Meissner & Lassiter, 2010; Snook, Eastwood, Stinson, Tedeschini, & House, 2010) have recently recommended that the guilt-presumptive and confrontational processes inherent in the Reid technique should be replaced by a noncoercive technique such as the PEACE model used in the United Kingdom. Importantly, the inherently deceptive and manipulative nature of the Reid technique (e.g., presentation of false evidence, minimization of the offense) raises important ethical/professional issues and concerns about its political legitimacy in a modern liberal society (Skerker, 2010). Skerker discusses three possible negative consequences of the Reid technique: risk of false confession, harm to police–community relations, and professional corruption of police interviewers. However, he does not consider these concerns “to be sufficient to rule out these interrogations on prudential grounds” (p. 114), provided interrogators are well trained and monitored and interviews are video recorded and corroborated.

Interviewing in the United Kingdom

The development of a formal police interview training approach in the United Kingdom began in the early 1990s following some landmark cases of miscarriage of justice involving confession evidence in which police interviewing and the courts came under close scrutiny and criticism (Gudjonsson, 2003). The authorities in England and Wales were quick to respond; they set up two Royal Commissions, which have initiated fundamental changes in laws and procedures, including those governing police interviews.

The PEACE model (PEACE is an acronym for Preparation and Planning, Engage and Explain, Account and Clarification, Closure, and Evaluation) has been in continuous use in the United Kingdom since 1993, when it was rolled out as part of a national training package for police officers. It is now being employed in many countries around the world, including New Zealand and Norway (Bull & Soukara, 2010). The PEACE model was developed out of sound psychological principles—following valuable collaborative work between academics, psychologists, police practitioners, and lawyers—and was intended to take into account vulnerabilities of some interviewees, with the aim to minimize the risk of false confession (Shawyer, Milne, & Bull, 2009). The initial planning and preparation stages require the interviewer to be well prepared for the interview, including possessing good knowledge about the case and the interviewee. The remaining phases involve an interview process, its closure, and its evaluation. The focus centers on fairness, openness, workability, accountability, and fact (truth) finding rather than merely obtaining a confession. Leading questions, heavy pressure, and psychological manipulation are avoided, thereby potentially reducing the risk of false confession while still producing true confessions (Shawyer et al., 2009).

Unlike their American counterparts, U.K. police interviewers are not allowed to lie to suspects or present them with false evidence in order to obtain a confession, further reducing the likelihood of a false confession (Kassin et al., 2010). In addition, the entire interview process in all interviews in England and Wales is electronically recorded. Initial reports on the PEACE model were positive, although the first major review of progress identified concerns in relation to communication skills, training and supervision, and the quality of interviews with witnesses (Clarke & Milne, 2001). To improve interview training and supervision, Griffiths and Milne (2006) recommend a “five- tier ” approach to reflect introductory, intermediary, and advanced interviewing skills training, as well as training in supervision and coordination of interviews in serious and complex cases.

Bull and Soukara (2010) point out that very few studies have actually investigated what happens during police interviews and that even fewer studies have investigated the relationship between police interview tactics and suspects' responses and behavior. Notable exceptions with regard to the latter are the studies by Leo (1996) in the United States and by us (Pearse & Gudjonsson, 1996, 1999) in the United Kingdom.

In an early study following the introduction of the PEACE model, we (Pearse & Gudjonsson, 1996) found no evidence of intrusive tactics or challenges by police; but in the later study of more serious criminal cases, we (Pearse & Gudjonsson, 1999) found that some police interviewers in the United Kingdom resorted to aspects of the American-style Reid technique in order to break down resistance in serious cases. The latter typically resulted in unreliable and in some instances false confessions (Gudjonsson, 2003). In more recent studies, Bull and Soukara (2010) discovered that police interviewers were generally complying with the PEACE model and that the manipulative and coercive techniques we (Pearse & Gudjonsson, 1999) identified were not present. However, Bull and Soukara found no clear relationship between the types of tactics used and confessions, although the researchers found that, unlike some other U.K. studies, confessions often did not occur at the beginning of the interview and that when confessions were made, there had been more confrontation and greater pressure in the police interview. The current authors are not aware of any reported real-life cases of false confession involving the PEACE model, but this has not been empirically investigated (Bull & Soukara, 2010). It is important to note that the PEACE model is still producing a high rate of confessions (Bull & Soukara, 2010; Gudjonsson, 2003).

Collaborative practice and research into police interviews continues in the United Kingdom. As a result, training for police officers has become more clearly defined, with a structure designed to better reflect an officer’s career development that is consistent with published national standards; new legislation includes specific witness categories with more specific procedures, making treatment of witnesses much more formal and better regulated; and since September 2001, there has been a noticeable increase in the numbers of terrorist suspects detained by U.K. police forces. One of the real tests of an effective model is its ability to evolve and adapt to its changing environment. For example, The Garda Síochána, the police service of the Irish Republic, has designed a model with the flexibility to adapt to the motivation and characteristics of interviewees rather than to their status as a witness, suspect, or victim. Thus, the interviewing officer needs to be adaptable enough to accommodate both cooperative and uncooperative individuals and vulnerable witnesses or suspects. This departure from the more traditional training approach is worthy of further review and evaluation. The PEACE model and the adapted Irish model enable the principles of the cognitive interview technique to be included for enhancing recollection among cooperative suspects (Fisher & Perez, 2007).

False confessions: The Reid technique versus the PEACE model

The extent to which the Reid technique and the PEACE model elicit false confessions is unknown. Since, unlike the Reid technique, the PEACE model is neither guilt presumptive nor overtly confrontational, it is widely assumed that it is less likely to elicit false confessions (Kassin et al., 2010; Meissner, Russano, & Narchet, 2010). Prior to the introduction of the PEACE model in the early 1990s in the United Kingdom, there were a number of cases of miscarriage of justice, although not all of these were due to coercive police interview tactics (Gudjonsson, 2010a). Any realistic comparisons between the two different techniques with regard to eliciting false confessions would need to consider the base rate of guilt among those interrogated. The higher the base rate of guilt of those interrogated, the lower the risk of false confession (Kassin & Gudjonsson, 2004). This kind of a comparison is easier to achieve from laboratory experiments than from field (real-life) studies in which the crucial variables (e.g., guilt, innocence, strength of the evidence against the suspect, suspect’s age) can be manipulated and controlled. Meissner, Russano, et al. (2010) present some emerging laboratory data showing that the inquisitorial U.K.-type interview technique is superior to the accusatory Reid-type technique in producing fewer false confessions and an increased number of true confessions when the CI technique of Fisher and Geiselman (1992) is used.

Of course, the limitation with all laboratory studies is that they cannot mimic the high stakes of real-life interviews, in which a number of highly complex factors interact—including fear and stress over involuntary confinement, the intensity and duration of the questioning, idiosyncratic vulnerabilities, the legal rights of the suspect in different countries, the presence and advice of lawyers, and the strength of the evidence against the suspect.