Want a Better Criminal Justice Response to Rape? Improve Police Interviews With Complainants and Suspects

The Complainant Interview

First and foremost in any rape investigation, there is the need to obtain an accurate and detailed account from the complainant. The hidden nature of offending means this account may be the only source of information for the investigation, no more so when the alleged offense occurs in a country where the suspect can exercise the right to silence. ² Memory research shows that how an account is elicited from an eyewitness is a system variable that can have a major influence on the accuracy and detail of information obtained (Geiselman et al., 1984; Lipton, 1977; Loftus & Palmer, 1974). As a special type of witness, applying these research findings to how information is elicited from an adult rape complainant can dramatically improve the quality of her account. Accurate detail is important in rape cases because of the increased reporting of offenses where the complainant and alleged offender are known to each other (Daly & Bouhours, 2010; Planty, Langton, Krebs, Berzofsky, & Smiley-McDonald, 2013). This means the issue of whether the complainant consented, or if the alleged offender had reasonable grounds to believe she consented, is likely to be central to establishing what occurred. For example, details about conversations between the complainant and the alleged offender and other ways the complainant may have expressed that she did not consent (e.g., that she was distressed or trying to physically resist him) are powerful evidential details.

The first of three main ways to ensure the complainant provides the most complete and accurate account is to create a social and cognitive environment that is conducive to the complainant recalling and reporting what happened. One vital social factor is to ensure that police attitudes encourage the complainant to stay in the criminal justice process when she initially reports to police, rather than police “filtering out” the complaint before a formal interview (Kelly, Lovett, & Regan, 2005). There is already a wide body of literature that examines the problems with disbelief about rape allegations and potential ways to improve police attitudes (Horney & Spohn, 1996; Jordan, 2001; Kelly et al., 2005). This response may include providing medical care and encouraging the complainant to participate in an interview, thereby formally entering into the criminal justice process. During the interview process, this includes establishing rapport with the complainant to create an environment that they feel at ease, listened to, and not judged (Fisher & Geiselman, 1992; Milne & Bull, 1999). Together, these factors may also make the complainant feel more comfortable and motivated to disclose highly personal but relevant information such as the precise nature of the sexual acts (Kebbell & Westera, 2011). Indeed, while specific aspects of a sex offense may be embarrassing and distressing (for both the complainant and the interviewing officer) this is exactly the information that is likely to be convincing to decision makers, and therefore interviewers should not shy away from eliciting this information. Some research into jury decision making indicates that “gruesome” violent evidence increases the likelihood juries will convict (Bright & Goodman-Delahunty, 2006, 2011). It is possible that this is also the case with gruesome sexual evidence. Despite these and other potential benefits to the investigation of making the complainant feel at ease, complainants often report feeling judged, especially when the alleged offender is known to them (Jordan, 2001; McMillan & Thomas, 2009; Patterson, 2011). Not surprisingly, from interviews with 20 adult complainants of sexual assault, Patterson (2011) found that when a complainant felt judged by the interviewing officer she was less likely to provide detailed information about the alleged offense. This is consistent with a great deal of research that suggests that the community is more likely to help the police, for example, by reporting crime and providing statements if they believe that they will be treated fairly and with respect (Tyler, 2001, 2003).

In addition to social factors, cognitive factors will influence what a complainant reports. Offenses may involve the use of alcohol or drugs and happen late at night. Arranging the interview at a time when the complainant is not tired or under the influence of alcohol or drugs is likely to improve her ability to concentrate and recall information from memory. A relaxed environment that is free from interruptions will also help the complainant concentrate to the level required to recall detailed information (Fisher & Geiselman, 1992). As well as providing a quiet and distraction-free interview room, this involves the interviewer allowing the witness to do most of the talking and not interrupt (see Fisher & Geiselman, 1992; Powell, Fisher, & Wright, 2005). Although there have been improvements in recent years, many police stations do not have private and comfortable rooms for interviewing complainant or other types of witnesses, and so interviews are conducted in a wholly inappropriate environment (Penney, 2007; Schollum, 2006).

Another important cognitive factor is to use questioning methods that encourage a narrative account from the complainant (Powell et al., 2005). Research-driven recommended practice is to avoid leading or suggestive questions that imply the desired answer and can lead to memory distortions (Fisher & Geiselman, 1992; Loftus & Palmer, 1974; Milne & Bull, 1999). Open questions or invitations that encourage unrestricted narrative responses are considered most desirable (Fisher & Geiselman, 1992; Milne & Bull, 1999; Powell et al., 2005). These types of questions allow a witness to control the flow of information and have been shown to generate more detailed, comprehensive, and accurate reporting (e.g., Fisher & Geiselman, 1992; Hutcheson, Baxter, Telfer, & Warden, 1995; Lipton, 1977; Powell et al., 2005). Open questions also enable the complainant to disclose potentially vital information that the interviewer may not have otherwise requested because they were not directly related to the reported crime–for example, that the alleged offender groomed or intimidated the complainant prior to the alleged offense (Kebbell & Westera, 2011; Westera, Kebbell, & Milne, 2013b). Rather than free narrative responses, police interviews are often dominated by interruptions and closed questions that require a short response (Clifford & George, 1996; Fisher, Geiselman, & Raymond, 1987; Snook, Luther, Quinlan, & Milne, 2012). Closed questions can be useful for clarifying aspects of the account or exploring ambiguous responses, but are unlikely to provide rich information and may result in decreased accuracy due to social compliance (Hutcheson et al., 1995; Lipton, 1977).

Using the cognitive interview (CI) is another way to improve the amount of information a complainant provides. This empirically validated method uses a number of memory aides, or mnemonics, that can increase the completeness of information remembered by an eyewitness (Fisher & Geiselman, 1992; Fisher, Milne & Bull, 2011; Geiselman et al., 1984). For example, a meta-analysis of the CI elicited an average of 41% more correct details when compared with control interviews, without compromising overall accuracy rates (Köhnken, Milne, Memon, & Bull, 1999; see Memon, Meissner, & Fraser, 2010 for a recent meta-analysis). The mnemonics used are derived from encoding specificity theory, which proposes that memory recall is aided by an increased overlap between encoding and recall conditions (Fisher & Geiselman, 1992; Geiselman, Fisher, Cohen, Holland, & Surtes, 1986; Tulving, 1972). For example, the mental reinstatement of context mnemonic involves the interviewer asking the witness to recreate the context of the crime by thinking back to where they were and how they were feeling at the time.

In addition to the benefits of an overall increase in recall, some CI mnemonics may be particularly useful in rape cases, such as reporting from different sensory modalities. Most witnesses will primarily report what they saw, so asking them to report everything that they heard, felt, tasted, and smelled may provide additional highly relevant information (Fisher & Geiselman, 1992). In rape cases, this may include specific details that go to the heart of whether sex was consensual–for instance, if the alleged offender knew that the sexual act was not consensual because the complainant explicitly said no, or it was obvious because she was crying and distressed. The CI has also been shown to produce additional information about thoughts and emotions when compared with a control interview (McCabe, Krauss, & Lieberman, 2010). This type of information may be vital for establishing issues around consent: for example, if the complainant froze due to fear, it is important that decision makers (e.g., prosecutors, jurors, judges) understand why she did not respond according to their stereotypes of how a victim should respond, a point we will explore further later.

While the interview methods described can enhance reporting, research indicates that officers often have difficulty applying the CI and using open questions (Clarke & Milne, 2001; Clifford & George, 1996; Griffiths, Milne, & Cherryman, 2011). Part of the difficulty is that traditional police training regimes do not contain many of the key elements required for learning skills (Ericsson, Krampe, & Tesch-Römer, 1993; Powell et al., 2005). In contrast to the short, intensive training sessions common in policing organizations (Schollum, 2006; Shepherd & Milne, 2006), developing interviewing skills requires training over multiple sessions, expert instruction and feedback, exemplars of good practice, and motivated participants (Powell et al., 2005). In England and Wales, Clarke and Milne (2001) suggested that training a small number of specialists could make this intensive training financially feasible for police. Despite this suggestion, New Zealand was the only country we could find that had published evidence of a police service adopting an ongoing development program with interviewers of adult complainants in a way that resembles evidence-based practice (New Zealand Police, 2012; this program has not been evaluated). Specialization may also assist with ensuring that police select interviewers who are genuinely motivated to improve their skills and have the resilience to deal with traumatized complainants.

In sum, the literature on interviewing complainants indicates that a conducive interview environment, and using well-validated methodologies of questioning and interviewing, can dramatically improve the quality of information elicited from a complainant. Later, we return to the issue of how this information can be captured and used in the criminal justice process. Using the methods outlined to obtain information from the complainant should also enable investigators to improve their planning for the other main aspect of the investigation: the suspect interview.

The Suspect Interview

How a suspect is interviewed is a system variable that is just as important as how the complainant is interviewed. If a guilty suspect confesses, then a conviction is almost inevitable; indeed, the suspect will usually plead guilty prior to trial, sparing the complainant the need to provide evidence in court (Gudjonsson, 2003). Of course, being suspected of a sex crime is unpleasant, whether one is guilty or innocent, and particular care must be taken to ensure the suspect’s rights are protected. Fortunately, an increasingly sophisticated literature indicates that there is much that can be done to ensure that suspects are treated fairly, while enhancing the likelihood of eliciting accurate accounts.

Most notably, the literature indicates that the evidence against a suspect is highly influential on suspects’ decisions to confess or deny. For example, Gudjonsson and Petursson (1991) surveyed 74 Icelandic prisoners concerning their reasons for confessing. Offenders were required to respond to questions using a Likert-type scale ranging from “not at all” (1 or 2) to “very much so” (6 or 7). The majority (55%) of offenders gave scores of 6 or 7 to the question: “Did you think the police would eventually prove you did it?” This was the most frequently given reason for confessing. Similarly, Moston, Stephenson, and Williamson (1992) investigated confessions for 1,067 suspects in the United Kingdom, who had been interviewed by detectives for a range of crimes. The results showed that when the evidence was rated weak by researchers, confessions occurred in less than 10% of cases. However, when evidence was rated strong by the researchers, confessions occurred in 67% of cases.

Specifically related to sex offenders, Kebbell, Alison, Hurren, and Mazerolle (2010) surveyed 43 convicted Australian sex offenders concerning how they were interviewed and looked at how this related to their decisions to confess or deny. Evidence-presenting strategies both influenced offenders’ decisions to confess to their own crimes and were suggested by respondents as strategies that investigating officers should use to increase the likelihood of a suspect responding honestly. The implication of this is that officers should ensure that they have the most evidence possible, and the previous section concerning the complainant gives some guidance as to how this can be achieved. Indeed, providing the enhanced account to the suspect and/or his lawyer has the potential to demonstrate the strength of the evidence, if it is strong, and may facilitate suspects’ truth telling.

Guilt may also affect decisions to confess. The previously mentioned research by Gudjonsson and Petursson (1991) indicated that 40% of respondents gave a “very much so” rating to the question: “Did you confess because you felt guilty about the offence.” However, in cases of sexual offending, while guilt may drive a desire to confess, shame and fear of exposure may also drive a desire to deny the offense and hide from the associated stigma. J. M. Read, Powell, Kebbell, and Milne (2009) therefore recommend adopting a non-judgmental approach that indicates to the suspect that his account will be heard, understood, and not judged. Convicted sex offenders strongly endorse interviewing strategies that they believe to be fair and not aggressive (Holmberg & Christianson, 2002; Kebbell et al., 2010). This is in contrast with confession-based approaches prevalent in many countries that start from the assumption that the suspect is guilty and even sometimes actively discourage the suspect from giving their version of events (e.g., Inbau, Reid, & Buckley, 1986).

Adopting a free narrative approach, as pointed out by J. M. Read et al. (2009), is another method that may encourage the suspect to give a detailed and accurate account–for example, asking the suspect (without assuming any wrongdoing) to give their version of the events or to explain the nature of their relationship with the alleged victim. This recommendation is consistent with a non-leading open-ended approach, which is the benchmark of best practice in prominent investigative interviewing protocols (e.g., Milne & Bull, 1999). It is also consistent with the approach adopted by many clinicians to encourage disclosure of sensitive information in therapeutic settings (Royal College of Psychiatrists, 2005).

The value of using broad invitations when initially raising the topic of concern is threefold. First, open-ended invitations are perceived as non-threatening and less interrogative when compared with a series of specific questions (i.e., those that dictate what specific information is required and offer the interviewee little time to collect his or her thoughts). This is because open-ended questions portray a relationship where interviewees’ perspectives are valued and heard, and these questions allow truly guilty suspects to portray themselves in a positive light, thus lowering their defenses. Second, open-ended questions minimize the risk that interviewers will raise specific case-related information (when discussing the allegation) in an attempt to confirm their preconceptions or prior knowledge (Meissner & Kassin, 2002; Nickerson, 1998; Wason, 1960). Third, this allows the suspects to put their own “gloss” on their confession if they confess. For example, child sex offenders may wish to emphasize their personal bonds and relationships (Benneworth, 2007), while rapists may focus on criticizing the alleged victim’s character and credibility. Both may wish to emphasize reasons why they are not fully culpable–for instance, that they are stressed or were under the influence of alcohol.

Investigation and Decisions to Charge

In the five countries examined in Daly and Bouhours’ (2010) review of attrition studies, ⁴ on average only 30% of all cases reported to police resulted in a decision to prosecute. Higher quality accounts from the complainant and alleged offender are likely to improve these prospects. More details about consent and what happened may also improve the evidential sufficiency of the case. Additional information may result in more investigative leads, which could increase the likelihood of corroboration with other evidence. For example, when there are multiple victims, more information about the alleged offender’s modus operandi may corroborate their accounts and could later be used as similar fact evidence.

The additional details from the complainant may also improve her believability and with it the prospect that the case results in prosecution. A “culture of suspicion” is one reason given for the low rate of prosecutions by police (Kelly et al., 2005, p. 51). In other words, the complainant is not believed due to tendency to overestimate the number of false complainants. Specifically, cases that depart from the common stereotype of “real rape,” when the offender is a stranger who physically overpowers a resisting complainant in a public place, are less likely to result in prosecution (Daly & Bouhours, 2010; Estrich, 1987). Research suggests that when these stereotypes are not met, the complainant may be seen as behaving counterintuitively or unreasonably and hence are blameworthy (Horney & Spohn, 1996; Jordan, 2001). This is particularly concerning because rape most commonly departs from this stereotype. For example, U.S. National Crime Victimization Survey data from 2005-2010 suggest that 78% of victims knew the offender and 55% of offenses happened at or near the victim’s home (Planty et al., 2013). It is important to note, however, that lawyers and members of the public have also been shown to hold these and other rape myths (Kerstetter, 1990; Schuller, McKimmie, Masser, & Klippenstine, 2010; Temkin & Krahé, 2008). Police, therefore, may choose not to prosecute if they anticipate there is no “reasonable prospect of conviction” due to the biases jurors hold.

The police gathering more complete information from the complainant about her thoughts and emotions may help overcome the negative effects of these “rape myths” (Ellison, 2007; Kebbell & Westera, 2011). This is because the information may make the complainant’s behavior more understandable to outside observers–for example, if the complainant explains that she did not physically resist the alleged offender because she was terrified and froze, or that she was concerned that resistance would cause the level of violence to escalate. Not only may officers find the account more believable, but it may also bolster their perceptions about how believable the account would be to a jury and hence increase the likelihood they charge the alleged offender. Access to a video record means that decision making by an investigator is open to scrutiny. An officer who holds strong rape myths and disbelieved the complainant may previously have “shut down” the case after the complainant interview and recorded “no complaint” or “insufficient evidence.” The ability to review the interview may reduce the likelihood that this will happen and may allow for effective case review and the monitoring of officer behavior.

Pretrial Decision Making

If the case results in a prosecution, a more complete and detailed record of the interviews is also likely to benefit pretrial decision making for both the prosecution and defense. Lawyers report that a video record enables them to improve case preparation because they can see firsthand what the complainant said and how they said it (Westera, Kebbell, & Milne, 2013a). For example, Sanders, Creaton, Bird, and Weber (1996) found that many witnesses with intellectual disabilities appeared far more competent in their written statements than in fact they actually were. Reviewing the interview may also help inform decisions to plea based on how credible a complainant or alleged offender is likely to appear to a jury. In addition, lawyers can scrutinize how the information was elicited to determine if this may have affected the reliability of that information (Heaton-Armstrong & Wolchover, 1992; Powell et al., 2005; Westera et al., 2013a). For example, information resulting from leading questions (e.g., “Did he put his fingers inside you that time?”) may be considered less reliable than that gained from open questions (e.g., “Tell me what happened.”).

The inclusion of additional information that may have previously been filtered out during the statement-taking process makes the process fairer and will also likely improve case preparation for both the prosecution and the defense. In the example we used earlier concerning a complainant reporting that she froze during the sexual assault because she was terrified, a prosecutor may decide to call expert evidence that this is a common response to being raped (Ellison, 2007). Or where the information may have a negative impact on juror judgments about credibility (e.g., that the complainant used recreational drugs on the night of the alleged offense), it may be more effective for a prosecutor to introduce this information in advance rather than waiting for the defense to do so in a more damaging way during cross-examination (Dolnik, Case, & Williams, 2003; Evans, 1995). A better understanding of the case may result in earlier guilty pleas or decisions to discontinue the prosecution in some cases. For the defense, when the prosecution case is strong, they may have a better basis from which to persuade the alleged offender to plead guilty. It may also reduce the likelihood a defendant will hope that the complainant will not turn up to give evidence or may be persuaded with intimidation to change their evidence.

For the benefits of improved accounts from the complainant and suspect to be realized, lawyers need to review the interviews before trial, which will take longer compared with reviewing a written statement (Westera et al., 2013a). Studies suggest that this seldom happens; instead, lawyers tend to rely on the often unreliable and always incomplete police statement or summary of interview (Baldwin & Bedward, 1991; HM Crown Prosecution Service Inspectorate & HM Inspectorate of Constabulary, 2007). Reviewing the interview is an opportunity to increase early resolutions and improve case outcomes, and is likely to save time in the long run. Next, we examine improvements that can be made to the criminal trial.

The Criminal Trial

The most powerful evidence in a rape trial is usually the account of the complainant. This is because, in an adversarial criminal trial, ⁵ the onus is on the prosecution to present the jury or judge with evidence that establishes the guilt of the accused beyond a reasonable doubt. The complainant is usually the main prosecution witness and traditionally gives their account of what happened in the courtroom in response to questions from the prosecutor (known as direct evidence). Defense counsel then cross-examines the complainant in an attempt to discredit her testimony. One system variable that may dramatically improve the quality of complainant evidence is to use her video-recorded police interview in place of live courtroom direct evidence.

There are three main reasons why the video may provide more complete and reliable information to a jury or judge. First, it may be months and often a year or years from the initial reporting of the alleged offending to the trial (e.g., Westera et al., 2013b). This time delay is likely to lead to memory decay that reduces the completeness of the information (Ebbinghaus, 1913; J. D. Read & Connolly, 2007; Rubin & Wenzel, 1996), and provides an opportunity for memory distortions that may reduce information accuracy (French, Garry, & Kazuo, 2008; Gabbert, Memon, & Allen, 2003; Loftus & Palmer, 1974; Tuckey & Brewer, 2003). Second, advocacy guidance recommends that prosecutors use questions that encourage short answers and control the witness. Studies of prosecutor questioning suggest that they predominantly use closed questions and, as previously discussed, this is likely to result in less detail than open questions that police are recommended to use (Kebbell, Deprez, & Wagstaff, 2003; Westera et al., 2013a; Zajac & Cannan, 2009). Third, the environment of the courtroom is likely to be detrimental to memory recall and reporting due to the stress of giving evidence (see Deffenbacher, Bornstein, Penrod, & McGorty, 2004; Yerkes & Dodson, 1908). The formal setting is also likely to discourage the disclosure of very personal details, which are often the highly relevant evidence, such as the sexual acts that occurred (Konradi, 1999).

To see what differences video evidence could make, Westera et al. (2013b) compared the content of transcripts of the complainant’s video-recorded police interview and their direct evidence in 10 actual rape cases in New Zealand. These transcripts were coded for details that went toward establishing consent (e.g., “I told him I didn’t want him near me”) and what happened during the commission of the alleged offense (e.g., from “He pushed me onto the bed” to “He rolled over and fell asleep”). The comparison found that details in interviews about these highly relevant matters were significantly more likely to be missing from, than included in, live direct evidence. Furthermore, details about cognitions suffered the greatest loss and were virtually non-existent in live evidence. As discussed, these types of details may be especially important in rape cases because they may provide a plausible explanation that helps counter rape myths. These findings suggest that video evidence could substantially increase the amount of evidence available to jurors or judges in rape cases.

Despite the potential benefits of video evidence, legislation enabling this method is only available in some countries ⁶ (Australian Law Reform Commission, 2010; Criminal Justice System, 2007; Mahoney, McDonald, Optican, & Tinsley, 2007), and in those countries, legal professionals tend to continue to rely on live evidence processes (Burton, Evans, & Sanders, 2006; Kingi & Jordan, 2009; Stern, 2010). This may be due in part to a traditional approach to courtroom practices, which assume that “an adult is perfectly capable of giving evidence” (Stern, 2010; Westera et al., 2013a). Over 40 years of memory research suggests this not the case (J. D. Read & Connolly, 2007). A reluctance to use video evidence has recently led to legal reforms in England and Wales, which require that the prosecution must apply for alternative ways of evidence (Home Office, 2011; McDonald & Tinsely, 2011).

More research is required to examine exactly how video evidence influences perceptions of adult complainant testimony. Some legal professionals are concerned that the video interview is less persuasive as evidence due to the video medium and narrative format of the interviews (Westera et al., 2013a). Video-recorded evidence however may also capture other features that make the complainant more persuasive, but are diminished in live evidence due to both the length of time passed and courtroom environment. For example, both police and prosecutors perceive that the complainant is often more emotional at the time of interview than at trial, and that this adds to complainant credibility (Westera et al., 2011, 2013a). These perceptions are supported by studies that found jurors rate rape complainants as less credible if they do not display emotion (Dahl et al., 2007; Rose, Nadler, & Clark, 2006; Wessel, Drevland, Eilersten, & Magnussen, 2006). Whatever the case, video evidence is likely to provide jurors with a more complete and fairer account of events.

A video-recorded account from the complainant may also remedy two system problems, created through traditional processes, that are used to discredit the complainant during cross-examination. First, when police produce a written statement, a complainant may be cross-examined on the contradictions or omissions between her live evidence and the statement. This is problematic because some of these inconsistencies are likely generated from the unreliable statement-taking process (Heaton-Armstrong & Wolchover, 1992). Second, when live evidence is given (regardless of how the police interview is recorded), a witness may be cross-examined on discrepancies between the interview and live evidence. These discrepancies may have occurred naturally due to memory decay (see J. D. Read & Connolly, 2007), or the effect of repeated recall attempts (see Brewer, Potter, Fisher, Bond, & Luszcz, 1999; Fisher, Brewer, & Mitchell, 2009). Of particular concern is that cross-examination is used not just to suggest that the witness is wrong about the discrepant item in question, but that their whole testimony is unreliable or dishonest (Evans, 1995). As jurors use consistency as a strong gauge of credibility, the process problems just described may undiagnostically diminish the credibility of the complainant’s entire testimony (Berman, Narby, & Cutler, 1995; Brewer et al., 1999; Fisher et al., 2009; Leippe & Romanczyk, 1989; Oeberst, 2012).

To recap, gathering more complete evidence from the complainant and suspect can assist decision makers in the investigation and prosecution process, but processes to ensure this happens are not fully utilized by the justice system. Now, we turn our attention to system variables that may serve to improve broader criminal justice system responses to rape and rapists. The quality of evidence will directly feed into justice outcomes around sentencing, alternative resolution, and intelligence.

Sentencing

With sentencing, a more thorough approach to interviewing the complainant and the suspect that gathers more details about the nature of the offending may help identify both mitigating and aggravating features of the crime as they were at the time, rather than relying on a self-serving and limited reinterpretation of events sometime later. A narrative approach to interviewing may also generate information about the offender’s lifestyle, mental health, alcohol and drug use. This type of information may help determine what sentencing principles apply: rehabilitation, incapacitation, punishment, or deterrence (Findlay, Odgers, & Yeo, 2009). For instance, a different response may be desirable when there is evidence of heterogeneous and prolific criminality compared with sexual fetishism or a single isolated sexual abuse incident. Sadistic behavior may indicate that a perpetrator needs to be incapacitated (Mann, Hanson, & Thornton, 2010). Having additional information that may help negate rape myths may also reduce complainant blame and increase the likelihood that sentencing decisions are informed by a fairer representation of the harm done. Indeed, appropriate sentencing may allow the offender to have access to treatment programs that not only reduce their risk, but may also give them the necessary skills to live happier, more productive lives (Ward & Brown, 2004). Additional problematic behaviors may also be identified; for instance, substance use at the time of the offense may indicate a general need for an alcohol or drug dependency intervention.

When rehabilitation is the objective, with the complainant’s consent, her interview record may be used as part of offender treatment. Viewing the complainant’s interview may help the offender understand the impact of their actions. It may also counter any justification and minimization of what happened, which is an important step in therapeutic interventions. Furthermore, having a full account may help identify, diagnose, and treat any mental health issues such as alcohol and drug dependence, sexual deviance, or mental disorder and prevent the offender from manipulating the treatment providers about what happened. There is also a growing desire to increase options for addressing sexual offending that are outside of traditional justice sector responses, such as restorative justice (Daly & Bouhours, 2010; McDonald & Tinsely, 2011). In these cases, where the complainant has initially reported the matter to police, the interview record may form a useful starting point to enable the alleged offender to understand the consequences of their actions.

Intelligence

We started this article by stating the low conviction rate for rape and have suggested a number of strategies that can be implemented to enhance how the police deal with rape. However, even with these strategies many guilty rapists will not be convicted. What should be done with these people? No criminal justice sanction can be applied without them being convicted of an offense. However, once the police are aware of serious allegations against an individual they should consider what risk that individual might pose in the future to the complainant and the community at large. While the previous sections have a substantial research basis what we are suggesting now concerns police intelligence and is more speculative, but has the potential to reduce sex offending and offending more generally.

Policing has traditionally been reactive: A crime is reported, and then it is investigated. More recently, some police services have moved to an intelligence-led approach, whereby police attempt to prevent crime before it occurs, or target emerging crime problems and prolific offenders (Ratcliffe, 2003). This is relevant to rape investigations as it is clear that many rape suspects have committed many other offenses in the past and will commit many offenses in the future. For instance, in a sample of 14,804 men in England and Wales convicted of contact sex offenses against an adult, overall 84% had previous non-sexual convictions with 76% having previous convictions for non-sexual violence and 71% for non-violent offenses (Howard, Barnett, & Mann, 2013). Of these offenders, 47% were subsequently convicted of a non-sexual offense after release, whereas only 2% were convicted of a contact offense against an adult (with all the previously mentioned difficulties around conviction, meaning that we must be cautious about how valid a measure this is of actual sex offending). What this means is that many offenders are not only rapists but are committing offenses across a range of domains. Indeed, high-risk sex offenders are more likely to commit a non-sexual offense than a sex offense as their next offense (see, for example, Smallbone & Rallings, 2013). This suggests that high-risk sex offenders may be at high risk of offending generally rather than just sex offending and, given the difficulty in prosecuting sex crime, prosecuting them for other offenses may be an alternative strategy.

One positive outcome of effective interviewing of complainants is that much more collateral information about suspects is available to investigators. For example, the authors recently reviewed 10 interviews with rape complainants for a study (Westera et al., 2013b). In these interviews, most of the complainants already knew the suspect and were familiar with his background. In the interviews, complainants gave a great deal of information about the suspects committing other offenses. For example, in one interview the complainant identified her alleged rapist as repeatedly driving drunk; in another, the complainant linked the suspect with a murder and named an accomplice; and in another, the complainant said the suspect had an illegal firearm and was dealing drugs. Clearly, this is intelligence information that can be used in police operations and can be combined with what the police already know about the suspect from his previous convictions and from other intelligence. While it may mean that the offender is not held accountable for the rape, he can be held accountable for other antisocial behavior.

Given the difficulties in successfully convicting rapists, it is important that intelligence is kept on those who have been suspected of rape. Previously, we have mentioned the skepticism that some police may have over some allegations. This skepticism will be reduced if multiple allegations are made against the same perpetrator by a number of women who do not know each other. An example of this is the case of serial rapist John Worboys, the London “Black Cab Rapist.” Worboys could possibly be England’s most prolific rapist and is thought to have raped more than 100 women. The sheer number of victims makes the case too complex to provide all the details here (for the inquiry, see Independent Police Complaints Commission, 2010), and there are multiple reasons why Worboys was free to continue raping for so long. However, three main factors stand out. First, many aspects of the investigation were of poor quality, especially the interviewing of complainants and the suspect. Second, some victims were not believed and so the allegations were not pursued. Third, there was no intelligence collation to show that there were a number of allegations against a taxi driver with the same modus operandi and a similar description. Of course, these errors compounded one another. However, if the intelligence had been available to investigating officers that linked the attacks it is likely they would have investigated more thoroughly, believed the complainants, and pursued the matter more tenaciously.

The implication of all this is that intelligence needs to be collated and used in rape cases. Indeed, by systematically using the information produced by rape investigations it might also be possible to generate intelligence around rape. A better understanding of a rapist’s particular modus operandi may help police to develop targeted prevention measures (Wortley & Smallbone, 2006). Similarly, allegations might be focused on particular times or places that may benefit from police attention. For example, offenses may be particularly likely in an entertainment district offering cheap alcohol or at particular times when lack of police presence may contribute to an increase in victimization.

Conclusion

The hidden nature of sexual offending and persistence of complainant-blaming attitudes make achieving just outcomes an ongoing challenge for the criminal justice system. For those involved in the process, the stakes are high, whether ensuring the recovery of survivors, preventing wrongful convictions, or minimizing the risk of reoffending. At a minimum, the criminal justice system must do all it can to ensure that the information central to determining outcomes in these difficult-to-investigate cases is gathered in a fair and transparent way. Key decision makers, both within and outside the criminal justice process, must have access to the most accurate and complete information possible. By applying the same standard of care expected of a crime scene examination to how information is gathered from the complainant and the alleged offender, investigators can improve the quality of information obtained and preserved. Methods for doing so are well documented, within control of the justice system, and can be implemented now. Adopting these methods will not only improve outcomes within and outside the justice process, but also the responsiveness of that process to those with the most at stake. Taken together, these approaches can improve rape investigation and prosecution dramatically.