Successful Criminal Prosecutions of Sex Trafficking and Sexual Abuse of Minors: A Comparative Analysis

Abstract

Despite increased awareness of sex trafficking of minors in the U.S., prosecution of traffickers remains difficult, in part because of victim uncooperativeness. There are questions about how that uncooperativeness is expressed, whether it is evident in successfully prosecuted cases, and whether it is unique to trafficked minors or it emerges in similar age victims of sexual abuse. To provide insight relevant to these questions, we compared appellate opinions in two types of successfully prosecuted criminal cases: sex trafficking and sexual abuse of adolescent victims. In the trafficking opinions, victims were rarely described as disclosing on their own or as knowing their trafficker before the victimization. The opinions also often alluded to the trafficking victims’ uncooperativeness and delinquency history, and frequently mentioned electronic evidence and prosecution experts. The sexual abuse opinions, in contrast, tended to suggest that victims’ own disclosures initiated the case, perpetrators were known and trusted adults, and caregiver support during the case was common. Finally, the sexual abuse opinions never explicitly mentioned victim uncooperativeness or electronic evidence and rarely mentioned expert testimony or delinquency. The different characterizations of the two case types highlight the need for improved education concerning effective prosecution of sex crimes against minors.

Introduction

Over the past several decades, attention to the problem of commercial sexual exploitation and sex trafficking, especially of minors, has grown dramatically (Victims of Trafficking and Violence Protection Act, 2000; Franchino-Olsen et al., 2022). This attention has highlighted the need to improve not only victim identification and intervention, but also prosecution of those who perpetrate this crime. Definitionally, sex trafficking involves the recruitment, transportation, transfer, harbor, or receipt of a minor via force, fraud, or deception for the purpose of engaging the minor in sexual acts in exchange for money or material goods (e.g., 18 U.S.C. §1591, 2022). Pandering, a related and often co-occurring crime, refers to the encouragement of a minor to become (or remain) a prostitute or the derivation of support from the earnings of an exploited minor knowing that the minor is a prostitute (e.g., Cal. Penal Code Section 266(i) (b), 2022). At times, trafficking and pandering also co-occur with sexual abuse. In such a case, the trafficker engages in sexual activity with youth while also exploiting them in exchange for earnings or material goods. Research on trafficking prosecutions has primarily surveyed professionals about their perceptions of victims and perceived challenges to prosecution (e.g., Brewer et al., 1997; Dianiska et al., 2023) or analyzed aggregate state or federal case-level criminal filings or outcome data (e.g., Farrell et al., 2016; Smith et al., 2009). These lines of inquiry, while valuable, do not provide insight into what trafficking cases actually look like in terms of their common characteristics or into whether trafficking case characteristics diverge from those in cases involving other sex crimes against minors, most notably child sexual abuse (CSA).

The goal of the present study was to begin to provide this insight by analyzing characteristics of successfully prosecuted trafficking cases involving youth victims. Specifically, we collected written appellate court opinions of criminal cases of trafficking or pandering of a minor and coded the opinions for details concerning the minor victims (e.g., background, indicators of cooperativeness), trafficker (e.g., manipulation strategies, relationship to the victim), and case itself (e.g., evidence presented in trial). Transcripts of trials can be obtained, but most are hundreds or thousands of pages in length and are costly and time consuming to collect. Written appellate opinions, in contrast, are more easily accessible, and they contain details regarding the reason for appeal but also summaries of the facts of the case. These summaries include details that extend well beyond those relevant to the appeal, such as information regarding how the case unfolded and why the guilty verdict was rendered. Analyzing these summaries thus provides valuable new insight into cases about which little is known.

To complement our analysis of appellate opinions of trafficking cases, we reviewed a second set of opinions comprised of successfully prosecuted CSA cases with victims similar in age to that of the trafficking victims (i.e., adolescents ages 12–17 years) (Kramer & Berg, 2003; Moore et al., 2020; Reid & Jones, 2011). As will be discussed, characteristics of both types of victims may make prosecutions difficult. Moreover, the two types of crimes are often heuristically grouped together as forms of sexual exploitation of youth, and, as mentioned, trafficking cases at times include charges of sexual abuse against the defendant. Whether the two types of crimes should be grouped together or whether unique features of each require their differentiation has not been examined. Further, whereas much is known about prosecuting CSA cases with younger (e.g., 3–12 years) victims (Cross et al., 2020), relatively little is known about prosecuting sexual abuse cases with adolescent victims. Our analytic approach afforded an opportunity to improve understanding of successful cases of adolescent sex abuse.

Victims’ History and Behavior

Several characteristics of victims of trafficking but also sexual abuse can affect the progression and outcome of a criminal trial (Lavoie et al., 2019; Nogalska et al., 2021). Some characteristics overlap between the two victim types, while others diverge. For example, sizeable numbers of both types of victims have histories of adversity and trauma exposure (Choi, 2015). This is especially true of trafficked youth, many of whom have previously suffered maltreatment (especially sexual abuse) and as a result experienced child protection and dependency court involvement, removal from home, and placement in out-of-home care. Also common among trafficked youth is having been homeless or having lived under conditions of extreme poverty (Fedina et al., 2019; Twis, 2020). These experiences, particularly in combination, increase youth’s risk for a range of negative outcomes, including health and academic problems, delinquency, criminal activity, trafficking, and prostitution (Chen et al., 2004; Conte & Von-Eden, 2017; Latzman et al., 2019). Some similar histories are evident in sexually abused adolescents. For instance, maltreatment in childhood (e.g., sexual abuse) is a risk factor for re-victimization in adolescence, and removal from home during childhood, placement in foster care, and experiencing repeated placement changes increase adolescents’ risk for sexual abuse (Finkelhor, 1983; Finkelhor et al., 2017; Mitchell et al., 2007; Raj et al., 2000).

Such histories, among both types of victimized youth, have implications for their potential cooperativeness and participation in prosecutions. Prior maltreatment and dependency court involvement, for instance, contribute to general feelings of mistrust of social services, the courts, and law enforcement (Heyman et al., 2020). Youth may hold police and social workers responsible for not protecting them from abuse, removing them from their parents’ care, or changing the youth’s placements. Youth may also feel no one listened to them during their case, a common complaint among foster youth (Goldfarb et al., 2021; Krinsky & Rodriquez, 2006), and may not believe that someone will listen to them about new victimization. Thus, even when professionals are available to help, youth with histories of trauma may be too skeptical, afraid, or distrustful of authorities to discuss more recent victimization.

Another characteristic commonly observed in trafficked youth is delinquent behavior. Many trafficked youth have histories of drug and alcohol use, gang membership, or repeatedly running away from foster or group homes (Choi, 2015; Fedina et al., 2019; Moore et al., 2020). Delinquent behavior often leads to encounters with the police, who may interrogate youth about their suspected criminal activity rather than question them about possible victimization (Halter, 2010). At times, these encounters result in victims disclosing having engaged in survival sex in exchange for food, shelter, or drugs while living on the streets. The victims may then be labeled prostitutes or be criminally prosecuted for prostitution (Halter, 2010; IOM & NRC, 2013). Alternatively, the victims may be detained in a juvenile detention facility, often on relatively minor charges (e.g., lying to an officer), as a way of protecting them from returning to a trafficker. Regardless, such encounters and experiences likely shape how the youth perceive law enforcement and likely how willing they are to cooperate later with those and related legal professionals (Dianiska et al., 2023). Of note, some adolescent victims of sexual abuse also engage in delinquent behavior (Edinburgh et al., 2013). However, their delinquent behavior generally occurs after the abuse and is believed to be a consequence of their victimization (Swanston et al., 2003). This is contrasted with trafficking victims, whose delinquent behavior tends to precede or co-occur with the trafficking (Franchino-Olsen, 2021).

Adolescent delinquent behavior combined with any resulting interactions with law enforcement could lead to several challenges to prosecuting traffickers and perpetrators of sexual abuse. For one, negative encounters with law enforcement (e.g., who may call youth prostitutes or detain them), as mentioned, reduce victims’ trust in authorities. Youth may be evasive and hostile when questioned by law enforcement officers, failing to believe that such professionals are there to help or protect the youth (Henderson et al., 2021). This hostility may reduce the officers’ provision of support during an interview and increase their use of interrogative practices to obtain information (Dianiska et al., 2023), both of which are associated with decreases in disclosure completeness in victims and suspects (Meissner et al., 2012; Saywitz et al., 2019). In addition, law enforcement and other professionals may not believe claims made by youth who appear complicit with the sexual activities, who behave in a hostile manner, or who have a history of delinquency (Luna et al., 2023; Winks et al., 2023). Finally, introducing details in court about victims’ delinquency could increase perceptions that they are responsible for their experiences, just as perceptions of adolescent victims’ responsibility for sexual assault increase when they are described as having engaged in age-inappropriate behavior (e.g., having been in a bar) (Rogers et al., 2016).

Victim Manipulation, Coercion, and Grooming

How trafficking victims are manipulated by their trafficker may affect their willingness to disclose to and cooperate with the authorities (Henderson et al., 2021; Nogalska et al., 2021), which in turn may affect the progression and outcome of a criminal case. The same may be true for sexual abuse victims. To date, relatively little attention has been paid in legal research to potential links among manipulation tactics used by traffickers and CSA perpetrators in cases involving adolescent victims, victim behavior, and prosecutorial decisions. Nor is it clear how often manipulation tactics represent core pieces of evidence in either type of criminal case, even though such tactics (e.g., use of force) are at times critical to the charges themselves.

Traffickers use a range of strategies to induce compliance. These include flattery or romance, which are often accompanied by implications that sexual contact between a minor and the trafficker is an expression of love. Other strategies include inducing feelings of fear or intimidation in victims, or creating feelings of obligation in victims by providing them with food, shelter, and protection (Reid, 2016). Law enforcement and other legal professionals often group traffickers into those who use seduction-focused or intimidation-focused tactics to manipulate victims (Eberhard et al., 2019; Merodio et al., 2020). Seduction-focused tactics involve a trafficker using affection, gifts, and feigned romance to pressure victims to engage in trafficking. Intimidation-focused tactics reflect a trafficker’s reliance on violence, threats, and physical coercion to induce victims into compliance. Common tactics used by CSA perpetrators include displays of affection, the provision of gifts, threats, showing victims pornography (e.g., to try to increase their interest in sexual activities), or describing sex as a way for the perpetrator to show he cares or as normal behavior (Beauregard et al., 2007; LeClerc et al., 2006; Whittle et al., 2014). CSA perpetrator tactics, though, have not been labeled as either seduction- versus intimidation-focused as they have with traffickers. Whether the two types (i.e., seduction or intimidation) of tactics are featured in successfully prosecuted cases of trafficking and sexual abuse is unknown but of interest given the role that manipulation plays in the charges themselves but also possibly with victim cooperativeness.

For instance, victims who believe that they and their trafficker are in love may be unwilling to disclose their experiences because these could implicate their romantic partner in wrongdoing. Victims may lie or falsely confess to a crime to protect their trafficker similar to tendencies observed in adolescent delinquents who at times falsely confess to protect a friend (Malloy et al., 2013). Adolescent victims of sexual abuse who have been manipulated by online perpetrators into sexual relationships may be uncooperative for similar reasons—to protect their perceived romantic partner. Victims of sex trafficking or sexual abuse who have been threatened or harmed may also be reluctant and ultimately uncooperative. Their reasons though may stem from fear of reprisal rather than from romantic feelings or a desire to protect their trafficker or perpetrator. However, if traffickers’ or perpetrators’ use of threats or harm led to physical evidence of control or abuse (e.g., threatening text messages, X-ray of a broken bone), prosecutors may not need to rely heavily on victims’ statements and testimony for a case to be successful. Overall, in light of the role that manipulation tactics may play in prosecutions of trafficking and sexual abuse, it is of interest to begin to explore how often and what types of tactics are mentioned in prosecuted cases, and whether and the frequency or type mentioned varies between trafficking and sex abuse cases.

Prosecution of Trafficking and Child Sexual Abuse

A few studies have examined criminal cases involving trafficking (e.g., Dianiska et al., 2023; Farrell et al., 2016; Smith et al., 2009), focusing on the type of evidence considered important to prosecutors and on the perceived importance of victims’ cooperativeness, testimony, and participation. Perhaps unsurprisingly, electronic and physical evidence (e.g., digital downloads, social media) are weighed very heavily by law enforcement when asked what types of evidence are most important to collect (Dianiska et al., 2023), and by prosecutors when deciding whether to file criminal charges (Farrell et al., 2016). These types of physical evidence, which may be perceived of as more objective than eyewitness or victim testimony, can be used to show control, payment, or coercion. Physical evidence is more likely when traffickers use physical or overt (e.g., threatening messages) intimidation rather than seductive (e.g., romance) manipulation tactics, and physical force, violence, and threats are associated with an increased likelihood of trafficking cases being prosecuted (Farrell et al., 2014).

In addition to prosecutors weighing physical evidence very heavily in their decision-making about moving forward with a case, they also seriously consider victims’ disclosures, cooperativeness, and availability to testify (Dianiska et al., 2023; Farrell et al., 2016; Reid, 2013; Smith et al., 2009). Prosecutors, for instance, express concerns that victims who believe that they are in a romantic relationship with their trafficker will be uncooperative and may fail to appear for court (Farrell et al., 2012). These concerns can reduce prosecutors’ willingness to accept cases when indicators of such a relationship are present (Farrell et al., 2012). Yet, victims may also fail to appear in court because they believe they need their trafficker for their own survival, they feel they need his financial assistance for material goods, or they are afraid of repercussions if they show, all of which could lead to behaviors (e.g., not showing for a hearing) suggestive of uncooperativeness (Nichols & Heil, 2015; Reid, 2013). How often trafficking victims, but also adolescent abuse victims, do not appear in court has yet to be examined.

It should be noted that victim cooperation and victim appearance are not identical. Both cooperative and uncooperative witnesses, including trafficking and CSA victims, may take part in interviews and testify. Cooperative victims may answer questions completely and disclose their experiences openly (Katz et al., 2021), whereas uncooperative victims may be evasive, unhelpful, or even hostile when asked about the abuse or sexual activities (Lindholm & Cederborg, 2016). Both cooperative and uncooperative responses have been documented in analyses of trafficking victims’ statements in forensic interviews and trials (Henderson et al., 2021; Nogalska et al., 2021), and law enforcement officers see evasiveness as one of their biggest challenges when interviewing trafficked minors (Dianiska et al., 2023). Prosecutors similarly cite a lack of victim cooperation as their biggest barrier to trafficking case prosecution (Farrell et al., 2016). Because uncooperative victims may still participate in interviews and trials, it is of interest to document how important both are in actual cases. Of note, however, cases with the most uncooperative victims or those who refuse to appear are unlikely to be accepted for prosecution in the first place. Thus, knowledge gleaned from successful cases is somewhat limited, although any mention of non-participation or uncooperativeness in successfully prosecuted cases, especially of trafficking but also CSA, provides some indication of their relevance separate from prosecutors’ perceptions.

Two other issues are important to consider when evaluating evidence, cooperation, and participation. First, although numerous studies have examined what types of evidence are important to CSA prosecutions, and how victim characteristics, including victims’ relationship to perpetrators, relate to case outcomes, few studies have focused specifically on adolescent victims. Yet, adolescent victims’ disclosure patterns, potential cooperativeness, and relationship to the perpetrator often diverge from those of child victims (Azzopardi et al., 2019; Goodman-Brown et al., 2003; Malloy et al., 2007) in ways that could affect cases’ progression and outcomes. And among studies that have examined the relations between how victims’ age and the likelihood of prosecution in CSA cases, findings have been mixed. Some studies, for example, have reported that cases involving younger victims are less likely to be prosecuted (Brewer 1997; Cross et al., 1994; Walsh et al., 2010), whereas other studies have reported that cases involving older victims are less likely to be prosecuted or that victim age is unrelated to the likelihood of prosecution (Bradshaw & Marks, 1990; Duron, 2018). However, victim age has varied substantially across studies, with some including infants through older adolescents; others excluding infants, toddlers, or adolescents altogether; and still others not reporting victims’ exact ages. In addition, in CSA cases with younger victims, factors such as maintaining a consistent disclosure, the victim taking the stand, and the presence of medical evidence all increase the likelihood of the case being accepted for prosecution and the likelihood of the prosecution being successful (Bradshaw & Marks, 1990; Brewer et al., 1997; Ernberg et al., 2018). Documenting whether similar factors (e.g., consistent disclosures, medical evidence) are evident in successful cases involving adolescent victims would provide insight relevant to important directions for follow-up investigations of CSA prosecutions.

Second, caregiver support is worth mentioning given the important role it plays in CSA prosecution success. Both Cross et al. (1994) and Goodman et al. (1992) found that caregiver support predicted greater prosecution likelihood in CSA criminal cases with victims spanning from the early preschool years through adolescence. Prosecutors often report that, without caregiver support, victims, whose testimony serves as a crucial piece of evidence, may not appear (Duron, 2018; see also Cross et al., 2020). Caregivers, for instance, may decide that a child participating in repeated interviews or taking the stand is simply too traumatic for the child to endure, leading to their refusal to help and then to charges potentially being dropped. A small study of adolescent victims of rape found a similar pattern: Victims who had received support from family or formal sources were more likely to participate in the prosecution than victims who had not received support (Feeney et al., 2018). We suspect that caregiver support is likely to be present in successfully prosecuted CSA cases with adolescent victims and as such may be could be mentioned in appellate opinions. Whether caregiver support is mentioned in successfully prosecuted trafficking cases is not clear. As already noted, trafficking victims often have histories of maltreatment and family violence, placement in foster care, and delinquency. These tendencies may be a function of poor caregiving or, at the very least, signal a strained relationship with caregivers. In either case, caregiver support of the adolescent, including during a criminal case, may be unlikely (Puigvert et al., 2021). Documenting whether caregiver support is mentioned in both types of successfully prosecuted cases would be of considerable interest and might suggest avenues for further research on methods of improving prosecution outcomes.

Present Study

The goal of the present study was to improve understanding of successfully prosecuted criminal cases involving charges of sex trafficking or pandering of a minor., and also of cases involving charges of CSA of similar age adolescents. To pursue this goal, we coded written appellate court opinions of trafficking and CSA cases for information pertaining to the victims, their relationship to the defendant, and other evidentiary details. Appellate opinions contain descriptions of key elements of the trial considered important for understanding the case history, progression, and decision. By analyzing these descriptions, we were able to gain unique insight into the most salient components of two types of poorly understood cases: those involving charges of trafficking and those involving charges of sexual abuse of adolescent victims.

Although our study was in many ways exploratory and descriptive, we were able to advance several tentative hypotheses based on extant literature. First, even though we expected delinquent behaviors to be mentioned relatively infrequently overall in successfully prosecuted cases, we anticipated that delinquency would be mentioned more often in trafficking than CSA cases. Second, we expected that descriptions of the defendant’s use of threats and coercion to manipulate victims would also be more common in trafficking cases than in CSA cases, as these manipulation tactics have been related to the likelihood of prosecution in trafficking cases (Farrell et al., 2016). It is unclear however, what other tactics would be mentioned, how these might differ between case types, or whether subsets of seduction- and intimidation-focused tactics could be identified. And third, few potential indicators of victim uncooperativeness were expected to be mentioned in both types of cases, given that the sample consisted of only cases that resulted in guilty verdicts. However, indicators of uncooperativeness were expected to be mentioned more often with the trafficking than CSA victims.

Method

Sample Selection

To obtain the legal opinions, we sent public record requests to District Attorneys’ Offices in 10 California (CA) counties between January and April 2020. California consistently ranks highest in the number of annual domestic trafficking cases (World Population Review Human Trafficking Statistics, 2022), and the state allows appeal as of right for all defendants found guilty at trial, making the state an ideal location from which to collect appellate opinion data. We selected the counties for inclusion based on their large population or because major transportation corridors pass through them. Our requests asked for defendant name, case number, and dates for all trials that took place between January of 2012 and December of 2018 which included charges of trafficking or pandering of minors (i.e., Cal. Penal Code Sections 236.1(c) (1),(c) (2); 266; 266h(b),(b) (1), (b) (2); 266i(b) (1), (b) (2); and 266(j)) and/or charges of CSA (i.e., Cal. Penal Code Sections 286 (c) (1), (c) (2) (B), (c) (2) (C); 288(a),(b); 288.5; 288.7; 288a; and 289(h),(i),(j)). Nine counties provided information. We reviewed the information to confirm trial dispositions and excluded cases that did not go to trial, failed to include one of the specified charges, or were consolidated (i.e., two cases involving the same defendant were combined into one; we kept only the consolidated case). A total of 1617 cases met our initial criteria. Eighty-four contained at least one charge of trafficking or pandering, and the remaining contained at least one charge of CSA.

Of the initial 1617 cases, 1365 ended in a guilty verdict on at least one charge. We then searched the 1365 cases to identify all of those that were contained in Westlaw, an online legal database (https://legal.thomsonreuters.com/en/products/westlaw) to identify those that had been appealed and for which a decision had been rendered. Westlaw publishes appellate courts’ written opinions, once available. In particular, all appeals are reviewed and decided by appellate court judicial panels that prepare a written summary of their decision. This summary or opinion contains the reason for appeal, the court’s decision and rationale, and a description of case details. The case details are not specific to the reason for appeal but instead concern facts of the case perceived by the panel as important to the case progression and outcome. The case detail description headings vary across appellate courts (e.g., they are labeled “Facts,” “Statement of Facts,” “Factual and Procedural Background,” “Summary of Evidence Presented at Trial,” “Factual Overview”), but virtually all appellate decisions contain descriptions.

In total, 1071 of the cases that contained a guilty verdict had been appealed and had written opinions available. Fifty-four included charges of trafficking (or pandering). Eleven of these opinions, though, were sting operations with no minors involved, and two of these opinions did not include any statement or description of case details. These 13 cases were excluded, leading to a final trafficking sample of N = 41, with victims ranging in ages from 10 to 17 years at the time of the trafficking. The remaining 1017 CSA appellate opinions were screened for whether a victim’s age was reported and for the total number of victims mentioned; 713 appellate opinions met these criteria. Of these, 221 reported that at least one victim fell between the ages of 11 and 17 years. From these, 60 cases were randomly selected with the restriction that (a) the number of CSA cases from each county was similar to the number of sex trafficking cases from the same county, and (b) the proportion of CSA cases with one versus more than one victim in each county was similar to the proportion of trafficking cases with one versus more than one victim in the same county. Unlike trafficking cases, several of which involved more than one defendant, only one CSA case included an adolescent victim with multiple defendants. We were thus unable to match on the number of defendants. Of the 60 CSA cases we selected, two did not actually go to trial. The defendant pled and later appealed his plea. These were removed, leading to a CSA sample of N = 58.

The final sample included 99 cases (N = 41 trafficking and N = 58 CSA). In total, cases contained 107 defendants, 161 minor victims, and 171 victim-defendant pairs (see Table 1). The court and appellate numbers for the final sample of cases (by defendant) are included in the Supplemental Materials.

Table 1.

Sample Characteristics and Evidence.

	Trafficking (N = 41)	CSA (N = 58)	Total (N = 99)
Case characteristics	N = 41	N = 58	N = 99
Single victim cases	24 (59%)	30 (52%)	54 (55%)
Single defendant	18 (44%)	29 (50%)	47 (48%)
Multiple defendants	6 (15%)	1 (2%)	7 (7%)
Multiple victim cases	17 (42%)	28 (48%)	43 (43%)
Single defendant	15 (37%)	28 (48%)	43 (43%)
Multiple defendants	2 (5%)	0 (0%)	2 (2%)
Number of defendants	N = 49	N = 58	N = 107
Male defendants	46 (94%)a	57 (98%)a	102 (95%)
Number of victims	N = 67	N = 94	N = 161
Female victims	63 (94%)	82 (87%)	145 (90%)
Male victims	3 (5%)	12 (13%)	15 (9%)
Transgender victims	1 (1%)	0 (0%)	1 (0.6%)
Number of victim-defendant pairs	N = 77	N = 94	N = 171
Types of evidence	N = 41	N = 58	N = 99
Prosecution evidence
Any victim testimony	30 (73%)	40 (69%)	70 (71%)
Eyewitness testimony	6 (15%)	9 (16%)	15 (15%)
Expert testimony	28 (68%)	21 (36%)***	49 (49%)
Electronic/social media evidence	32 (78%)	16 (28%)***	48 (48%)
Physical evidence	22 (54%)	11 (19%)***	33 (33%)
Defendant negative character	20 (49%)	12 (21%)**	32 (32%)
Pre-text/sting calls	3 (7%)	10 (17%)	13 (13%)
Defendant confession	6 (15%)	16 (28%)	22 (22%)
Medical evidence	2 (5%)	14 (24%)*	16 (16%)
Defense evidence
Any defendant testimony	10 (24%)	27 (47%)*	37 (37%)
Eyewitness testimony	0 (0%)	0 (0%)	0 (0%)
Expert testimony	5 (12%)	7 (12%)	13 (12%)
Positive defendant character	4 (10%)	8 (14%)	10 (11%)
Mean prosecution evidence types mentioned (range)	3.61 (0–7)	2.40 (0–5)***	2.85 (0–7)
Mean defense evidence types mentioned (range)	.46 (0–3)	.71 (0–3)*	.60 (0–3)

Note. Ns and percentages are reported for the evidence type to reflect specific case characteristics (e.g., single v. multiple victim, etc.) relative to the total number of cases. Asterisks reflect whether results were significant at p < .05 (*), p < .01 (**), and p < .001 (***) levels comparing trafficking and CSA cases.

^aThe three non-male trafficking defendants were female, and the one non-male CSA defendant was transgender.

Appeal Opinions Coding

Senior authors on the paper with expertise in child maltreatment and the law developed a list of details to be extracted and coded from the opinions. A team of seven graduate and undergraduate research assistants was then trained to extract and code relevant information separately for each victim-defendant pair. Reliability was established on 18% of the victim-defendant pairs using Gwet’s AC₁. This approach was preferred over Cohen’s Kappa because several of the coded pieces of information (e.g., mention of caregiver support) were judged disproportionally more absent than present across coders (Cohen’s Kappa can be misleading in situations where agreement is not homogenous across judgments and it does not account for agreement based on chance ratings; Gwet, 2008, 2014). Reliability was adequate across coders (M = .80, range = .66–.91 across the 31 calculations; see Tables for variables).

Coded information was classified into one of four categories: case and victim background, victim-defendant relationship dynamics, victim contact and interactions with legal authorities, and other case evidence. Most information was located in the case detail sections of the opinion (e.g., “facts of the case”), which usually ranged from 1.5 to 4 pages in length. However, the entire opinion (usually 4 to 16 pages) was reviewed for codable information.

First, potentially relevant information was copied and pasted into Excel. Second, numeric codes were applied to the extracted information to capture all characteristics of interest. The codes were entered into a separate variable database. Coded characteristics varied in terms of the level at which they were coded. Some characteristics were scored at the case level (e.g., whether an expert testified in the case). Others were scored at the victim level (e.g., each victim’s age when the abuse occurred and when they took the stand), at the defendant level (e.g., defendant gender), or at the victim-defendant relationship level (e.g., threats to obtain compliance). The latter is particularly important given that defendants could have had different relationships with each victim (e.g., one victim might be a blood relative and another victim a friend of the first), used different types of manipulation tactics (e.g., threats with one victim, but enticements with another), and even varied in whether or not they had sexual contact with each victim. All codes applied both to trafficking and CSA cases.

Also, for many of the coded characteristics, we distinguished among whether the opinion explicitly stated a characteristic was present (e.g., victim testified), explicitly stated a characteristic was absent (e.g., victim refused to testify), or did not mention the characteristic (e.g., no mention of victim testimony). In doing so, we were able to distinguish features of the case deemed sufficiently important by the appellate panel to warrant their inclusion in the opinion from features not considered sufficiently key to warrant their direct mention, whether they occurred or not in the actual trial.

Case and Victim Background Details.

Basic case information included the number of victims mentioned, number of defendants in the case, and gender of each. The age when the victimization began was documented (all victims had this age included, as it was a key matching criterion). Age when the abuse ended was documented if it was mentioned or it was coded as not mentioned. Victim testimony was coded as testified, did not testify, or not mentioned. For victims who testified, their age and the type of hearing (preliminary hearing, trial) were documented when possible, or these characteristics were coded as not mentioned. Childhood exposure to adversity was coded as mentioned or as not mentioned. Examples of adversity included comments about a victim having been in the dependency system or homeless. Psychological problems, learning disabilities (e.g., a victim’s “Individual Education Plan”), and delinquent or high-risk behaviors were each coded as mentioned or not mentioned. When delinquency or high-risk behaviors were mentioned, we then coded for type, separating running away, drug or alcohol use (excluding when forced), theft/robbery, stripping, prostitution unrelated to the current case, lying (e.g., the victim lied about her age to the defendant), and other (e.g., the victim had been previously detained).

Victim-Defendant Relationship Dynamics and Manipulation Tactics.

Details concerning the victim and defendant’s relationship and the defendant’s manipulation strategies were coded separately for each victim-defendant pair. The type of relationship between the defendant and each victim prior to the start of the victimization was coded as one of the following: (1) trusted caregiver (e.g., family member, parent’s romantic partner); (2) friend or acquaintance (e.g., “boyfriend she met at school”); (3) stranger; or (4) no mention of the defendant’s relationship or unclear relationship to the victim.

Whether the opinion explicitly mentioned sexual contact with the defendant (as opposed, for instance, in trafficking cases, with a procurer) was scored as: (1) explicitly stated contact occurred, (2) explicitly stated contact did not occur (it was attempted or no contact), or (3) did not mention whether contact occurred. Follow-up codes for contact distinguished indications of consensual or non-consensual contact, when these were present. These were coded separately because both might have occurred on separate occasions. Examples of consensual sexual contact included mentioning that the defendant and victim had intercourse on several occasions, the “victim seduced” the defendant, or the “victim wanted” the sexual activity. Examples of non-consensual contact included mentioning force, the victim trying to resist, or the victim described as “having given up resisting because it was unsuccessful.”

We coded whether the opinion included information on manipulation tactics employed by the defendant (mentioned versus did not mention). When mentioned, we coded each type separately: physically intimidating the victim (e.g., threats), isolating the victim, providing enticements (e.g., gifts) to the victim, psychologically manipulating the victim (e.g., he said he was “touching out of love”), providing drugs or alcohol to the victim to reduce inhibitions, normalizing sexual activity (e.g., he said it is “OK to have sex at this age”), exposing the victim to pornography, requiring the victim to work off debts (e.g., rent), and indicating romantic overtures (e.g., the defendant was the victim’s boyfriend).

Victim Cooperativeness.

Other codes captured details relevant to how cooperative the victim was with the authorities, including social services, law enforcement, and the courts. How the victimization was discovered, when mentioned, was coded into one of three mutually exclusive categories: (1) victim-initiated (e.g., victim contacted the authorities), (2) other-initiated (e.g., report from a family member, another victim disclosed), or (3) law enforcement-initiated (e.g., victim was caught in a compromising situation or during a sting operation). Comments about victim cooperativeness were coded as present or not, with comments being further coded into: (1) consistently indicate uncooperative (e.g., victim was hostile, victim ran away from law enforcement officer, victim lied about her age to the authorities), (2) inconsistent comments about cooperativeness (e.g., victim initially lied about her relationship, then described her trafficking experiences, and later recanted), or (3) comments implying consistent or always cooperative (e.g., victim disclosed full details and identified other victims).

Evidence.

Forms of evidence were coded as mentioned or not mentioned. These included electronic evidence (e.g., phone records, text messages, and social media posts), physical evidence (e.g., cash, condoms), medical evidence (e.g. pregnancy test), defendant testimony, victim testimony, defendant admission or confession, other witness testimony for the prosecution, other witness testimony for the defense, pretext calls (when law enforcement records a call between a victim or confederate and defendant to elicit incriminating admissions), expert testimony for the prosecution (e.g., police describing trafficker behaviors), and expert testimony for the defense (e.g., psychologist describing differences between perpetrator characteristics and the defendant’s characteristics). We also distinguished whether each mentioned piece of evidence was prosecution-focused (e.g., electronic/social media posts, physical, or medical evidence) versus defense-focused (e.g., positive defendant character).

Results

Results are presented in four sections: general case characteristics, victim characteristics, the dynamics of the victim-perpetrator relationship, and evidence. Within each section, we compared trends between the trafficking and CSA cases. Also, within and across sections, the analyses were conducted at different levels depending on variables of interest (e.g., case level, victim level). Most analyses consisted of Pearson χ² tests (Fisher’s exact tests when Pearson χ² assumptions were not met) and t-tests. In the text and Tables, the Ns generally refer to the number of characteristics mentioned versus not mentioned. When subcategories among the mentioned characteristics were coded, percentages reflect the subcategory frequency out of the number mentioned.

Case Characteristics

Descriptive data on the two case types are presented in Table 1. In nearly half of the cases (42% of trafficking and 48% of CSA), more than one victim was mentioned. In the trafficking cases, the number of victims mentioned ranged from one to seven. In the CSA cases, the number of victims mentioned ranged from one to five. A few trafficking victims were adults at the time of the trafficking, whereas several CSA victims were under the age of 10 years at the time of the abuse. All victims were included in counts of the total number of victims mentioned, but those outside of our target age range (10–17 when the trafficking began or 11–17 when the abuse began) were not included in subsequent analyses of the victim characteristics.

About one-fifth of the trafficking cases included multiple defendants, and 3 defendants were female. Only one CSA case had more than one defendant, and one CSA defendant was transgender (none was female). Finally, all 107 defendants were found guilty of one or more felonies. No case was fully overturned on appeal. All 49 trafficking defendants were found guilty of at least one charge of trafficking or pandering. Thirteen trafficking defendants were also charged with CSA, and 12 were found guilty of CSA in addition to being found guilty of trafficking or pandering. Of the 58 CSA defendants, 57 were found guilty of CSA, and one was instead found guilty of forcible rape (PC 261(a)) against another victim who was 18 at the time but acquitted of the CSA charge.

Victim Characteristics

Analyses concerning characteristics of the victims themselves (N = 161, 67 trafficking, 94 CSA) revealed some similarities between the two types of victims but also important differences, especially in terms of the victims’ presentation and risk for uncooperativeness. Considering basic demographics, nearly all victims were female, although 5% of the trafficking and 13% of the CSA victims were male (one transgender trafficking victim was mentioned; Table 1). In addition, even though we selected CSA cases whose victims closely matched the ages of the victims in the trafficking cases, on average, trafficking victims were still significantly older when the victimization began, t (157) = 8.15, p < .01, and ended, t (134) = 6.33, p = .01 (Table 2). The difference between the mean ages when the victimization ended, though, was not as large as the difference between the mean ages when the abuse began. This trend suggests that the duration of the victimization in successfully prosecuted cases may be shorter for trafficking than CSA situations, a possibility worth exploring in the future. We should note, though, that too few opinions explicitly mention duration to allow for direct comparisons.

Table 2.

Victim Characteristics.

	Trafficking (N = 67)	CSA (N = 94)	Total (N = 161)
Demographic details
Age when victimization began	15.35 (N = 65)	13.29 (N = 94)**	14.14 (N = 159)
Age when victimization ended	15.51 (N = 49)	14.74 (N = 87)*	15.01 (N = 136)
Legal involvement
No mention of victimization discovery	n = 21	n = 30	n = 51
Mentioned victimization discovery	n = 46	n = 64	n = 110
Victim-initiated disclosure	10 (22%)	47 (73%)***	57 (52%)
Law enforcement report (e.g., via sting)	30 (65%)	3 (5%)	33 (30%)
Another adult reported suspicions	6 (13%)	14 (22%)**	20 (18%)
No mention of victim cooperativeness	n = 34	n = 57	n = 91
Mentioned victim cooperativeness	n = 33	n = 37	n = 70
Victim described as cooperative	10 (30%)	29 (78%)*	39 (56%)
Victim described as uncooperative	3 (9%)	1 (3%)	4 (6%)
Victim described inconsistently	20 (61%)	7 (19%)***	27 (39%)
No mention of caregivers	n = 62	n = 51	n = 113
Mentioned caregiver support	n = 5	n = 43	n = 48
Caregiver was supportivea	5 (100%)	37 (86%)***	42 (88%)
Caregiver was unsupportivea	0 (0%)	10 (23%)	10 (21%)
No mention of victim testimony in trial	n = 20	n = 32	n = 52
Mentioned victim testimony in trial	n = 47	n = 62	n = 109
Victim testified in trial	40 (85%)	62 (100%)	102 (94%)
Victim did not testify	7 (15%)	0 (0%)	7 (6%)
Mean age when victim testified	15.93 (n = 27)	21.56 (n = 39)***	19.29 (n = 66)
Background and history
No delinquency mentioned	n = 23	n = 74	n = 97
Delinquency mentioned	n = 44	n = 20 ***	n = 64
Mean behaviors mentioned (range)	1.40 (0–5)	.33 (0–3)	.75 (0–5)
Types of delinquency mentioned
Drug use	11 (16%)	7 (7%)	18 (11%)
Prior prostitution	30 (45%)	6 (6%)***	36 (22%)
Running away	25 (37%)	3 (3%)***	28 (17%)
Alcohol consumption	2 (3%)	4 (4%)	6 (4%)
Stealing/robbery	2 (3%)	3 (3%)	5 (3%)
Lying	9 (13%)	3 (3%)**	12 (7%)
Sexual promiscuity/stripping	4 (6%)	2 (2%)	6 (4%)
Other (e.g. juvenile probation)	11 (16%)	4 (4%)**	15 (9%)
No adversity mentioned	n = 49	n = 84	n = 133
Prior adversity exposure mentioned	n = 18	n = 10**	n = 28
No emotional/academic problems mentioned	n = 63	n = 83	n = 146
Emotional/academic problems mentioned	n = 4	n = 11	n = 15

Note. Ns and percentages are reported for variables to reflect how frequently the specific variables were indicated in relation to the number of times the broader characteristics were mentioned. Asterisks reflect whether results were significant at p < .05 (*), p < .01 (**), and p < .001 (***) levels comparing trafficking and CSA victims.

^aFour caregivers were described as both supportive and unsupportive and are included in both counts.

Of greater interest, though, was how other victim characteristics differed between the two case types, including characteristics potentially relevant to victim cooperativeness. One such characteristic concerned how the victimization was discovered, which was often mentioned in the opinions. Comparisons between the trafficking and CSA cases suggested that the victims were discovered or identified via very different processes, χ² (2) = 47.64, p < .001. Whereas very few trafficking victims’ own disclosures initiated the criminal case, a large majority of CSA victims’ disclosures did so (Table 2). Identification of the trafficking victims instead tended to occur as a result of law enforcement intervention, either while officers were patrolling specific locations or during sting operations. Only three CSA victims were first identified by law enforcement. For the remainder of both trafficking and CSA victims, the courts mentioned indirect methods of victimization discovery, which often involved concerned family members reporting their suspicions.

Some opinions explicitly mentioned that victims were cooperative. This occurred significantly more often in reference to the CSA than trafficking victims, χ² (2) = 16.34, p < .001. When trafficking opinions mentioned cooperativeness, they tended to describe victim behaviors that suggested uncooperativeness or varying levels of cooperativeness over time (Table 2).

Other noteworthy characteristics that may be related to victim cooperativeness include whether victims testified and whether the victims’ caregivers were mentioned. Two-thirds of the opinions explicitly mentioned the victims’ testimony in trial, with no differences between the two case types in how often this occurred. When opinions mentioned the victims’ ages at the time they testified, trafficking victims were significantly younger than CSA victims, t (64) = 4.46, p < .001 (Table 2). Although difficult to interpret due to many cases not mentioning whether victims testified or the victim’s age when this occurred, this difference suggests an important direction for future research, namely, how case length may vary across different types of cases involving sex crimes against minors. Also, about one-third of the opinions explicitly mentioned caregivers, although this was primarily in the CSA cases and rarely in the trafficking cases, Fisher’s exact test = 29.54, p < .001. Statistical comparisons between cases in how caregivers were described were not possible due to so few trafficking victims’ caregivers being mentioned. None-theless, it is worth noting that, in the CSA cases, caregivers were nearly always described as supportive of the victim (Table 2). In 10 CSA opinions, the opinions also made a comment that caregivers were non-supportive, although in four of these, opinions mentioned both supportive and non-supportive behaviors. Overall, though, our finding that caregivers were mentioned more often in legal opinions about CSA than trafficking suggests that caregivers may well have played a more visible role in how the CSA trials unfolded.

Some opinions also described other forms of presenting victims’ statements at trial. For instance, a small number of opinions (n = 15) described victims’ testimony from a preliminary hearing. This occurred slightly but non-significantly more often with trafficking (n = 11) than CSA (n = 4) victims (Fisher’s exact, p > .05). Pre-trial interviews were mentioned in reference to just over one-third of the victims (22 trafficking and 34 CSA victims), with no significant differences emerging in how frequently these were mentioned about trafficking versus CSA victims. Finally, opinions infrequently referenced hearsay, either in the form of victims’ recorded statements being presented at trial (7 trafficking, 4 CSA) or someone else describing victims’ statements (6 trafficking, 5 CSA). When both types of hearsay were combined, hearsay was mentioned significantly more often with trafficking than CSA victims, χ² (1) = 4.09, p = .043. If trafficking victims are less cooperative and hence at greater risk of being unavailable during the trial than CSA victims, having a record of their statements (either via a preliminary hearing or recorded interview) can be valuable prosecutorial evidence in the case.

Our final victim-level analyses focused on delinquent behavior; prior adversity; and emotional, learning, or other academic problems. How often the two types of opinions mentioned these characteristics was of interest in part because the characteristics could be linked to victim cooperativeness but also because differences in the characteristics being mentioned between case types could illustrate important variations in how the victims were being portrayed at trial. First, opinions mentioned prior delinquent activity significantly more often when discussing trafficking victims, χ² (1) = 32.19, p < .001. Opinions also mentioned a greater number of different types of delinquent behaviors committed by the trafficking victims (0–5 types mentioned) than the CSA victims (0–3 types mentioned), t (159) = 6.44, p < .001 (see Table 2). When we examined which delinquent behaviors were mentioned most often, prior prostitution was most common followed by runaway behavior and then drug use. As might be expected, prostitution and runaway behavior were mentioned significantly more often in regard to the trafficking than the CSAvictims, as was the victim having lied in the past, χ²s (1) >7.26, p < .01. The difference between victims in drug use approached significance (p = .08).

Second, a small number of opinions mentioned victims’ exposure to trauma or adversity (Table 2), more often when discussing the trafficking than CSA victims, χ² (1) = 7.17, p = .007. And third, emotional, behavioral, and learning problems were mentioned very infrequently, with no significant differences emerging when discussing the two types of victims (Table 2). Overall, therefore, appellate opinions characterized trafficking more than CSA victims as troubled youth who engage in risky behavior with difficult pasts.

Victim-Defendant Relationship

A unique aspect of our investigation concerned our analysis of the dynamics of the relationship between the victim and defendant. Because relationship type and manipulation strategies could vary across victims and defendants, analyses were conducted at the victim-defendant pair level (N = 171, 77 trafficking, 94 CSA).

First, we focused on how the victim and defendant knew each other, categorizing the defendants’ relationships to victims as one of the following: strangers prior to the victimization, trusted adult caregivers (e.g., adult family members/relatives, parent’s romantic partner), or friends/acquaintances. When we compared victims’ and defendants’ relationship between the two case types, substantial differences emerged, X² (2) = 56.76, p < .001. As shown in Table 3, trafficking defendants were far more likely to be strangers prior to the victimization than were CSA defendants. The latter tended to be trusted caregivers (only one trafficker was a caregiver). Thus, commercial exploitation by a family member was uncommon in our sample of successfully prosecuted trafficking cases. Finally, defendants were acquaintances or friends in few trafficking cases, but many more CSA cases.

Table 3.

Characteristics of Victim-Defendant Relationship and Defendant’s Manipulation Strategies.

	Trafficking (N = 77)	CSA (N = 94)	Total (N = 171)
Type of relationship
None mentioned	n = 9	n = 5	n = 14
Type mentioned	n = 68	n = 89	n = 157
Stranger	56 (82%)	21 (24%)***	77 (49%)
Relative/caregiver	1 (2%)	33 (37%)***	34 (22%)
Acquaintance	11 (16%)	35 (39%)***	46 (27%)
Type of sexual contact
None mentioned	n = 35	n = 3	n = 38
Explicitly stated No contact occurred	n = 6	n = 1	n = 7
Explicitly stated contact occurred	n = 36	n = 90***	n = 126
Type of contact explicitly noted	n = 19	n = 85	n = 104
Indicated consensual at some point	7 (37%)	6 (7%)	13 (12%)
Indicated nonconsensual at some point	12 (63%)	79 (93%)***	91 (88%)
Type of manipulation strategies employed
No strategies mentioned	n = 3	n = 20	n = 23
Multiple strategies mentioned	n = 27	n = 29	n = 56
Mean strategies mentioned (range)	1.34 (0–5)	1.10 (0–7)	1.21 (0–7)
Strategies mentioned	n = 74	n = 74	n = 148
Enticements	10 (14%)	13 (18%)	27 (18%)
Isolation	10 (14%)	17 (23%)	28 (19%)
Drug use	9 (12%)	6 (8%)	16 (11%)
Psychological coercion	14 (19%)	16 (22%)	33 (22%)
Threats or physical force	29 (39%)	21 (28%)	50 (34%)
Financial con or control	2 (3%)	0 (0%)	2 (1%)
Presents sex as normal	11 (15%)	12 (16%)	23 (16%)
Showing pornography	0 (0%)	9 (12%)	9 (6%)
Romantic overtures	18 (24%)	8 (14%)*	28 (19%)

Note. Ns and percentages are reported for subsets of variables to reflect how frequently the specific variables were indicated in relation to the number of times the broader characteristics were mentioned. Percentages add up to more than 100% because multiple strategies could be present in cases. Asterisks reflect whether results were significant at p < .05 (*), p < .01 (**), and p < .001 (***) levels comparing trafficking and CSA victim-defendant relationship.

Second, we considered the type of sexual activity mentioned. Opinions explicitly mentioned sexual contact in about three-fourths of the cases, including nearly all CSA and about half of the trafficking cases, Fisher’s exact = 55.97, p < .001. Opinions directly stated that no sex occurred between the victim and defendant in a few cases, only one of which involved CSA (sexual contact was pursued via texting and phone calls but discovered via a sting operation prior to its actual occurrence). Also, when the appellate opinions mentioned the type of contact, they were more likely to describe it as consensual with the trafficking than with the CSA victims. With the CSA victims, the appellate opinions most often explicitly described the contact as non-consensual, χ² (1) = 79.68, p < .001.

Our third set of analyses focused on how defendants induced compliance in victims. Manipulation strategies were mentioned in nearly all of the trafficking opinions (74 out of 77 trafficking victim-defendant pairs), a perhaps unsurprising finding given that manipulation and coercion are central to the trafficking charge itself. However, manipulation strategies were also mentioned quite often in the CSA opinions (74 out of 94 CSA victim-defendant pairs), suggesting that strategies defendants used to induce compliance may have been salient aspects of CSA trials as well. When we examined the types of manipulation tactics mentioned, similar trends emerged between the trafficking and CSA cases. The most common tactics mentioned were physical threats or force, psychological manipulation, and isolation, followed by normalizing sex, enticements, and romantic overtures (Table 3). Less common were drugging the victim, exposing the victim to pornography, and financial con/debt bondage. Only one tactic, romantic overtures (e.g., the victim or defendant referring to each other as boyfriend/girlfriend), was differentially mentioned between case type, being mentioned significantly more often as a tactic employed by trafficking than CSA defendants, χ² (1) = 4.66, p = .031.

Most opinions only mentioned one type of manipulation tactic used by the defendant, and no differences were evident in the percent of trafficking (35%) versus CSA (31%) opinions that mentioned a defendant using more than one tactic, χ² (1) = 0.34, p = .55. When we counted the number of tactics mentioned for each defendant-victim pair, the mean was slightly but non-significantly higher for the trafficking defendants than CSA defendants (M = 1.34, range = 0 to 5, and M = 1.11, range 0–7, respectively) t (169) = 1.16, p = .25.

Trial Characteristics

Our final analyses focused on the trial itself, specifically on the types of evidence mentioned. Such analyses are relevant to prosecutorial decisions about whether to move forward with a case, given that prosecutors often express concerns about evidence, especially when they believe their case relies heavily on victim cooperation and testimony. The forms of evidence mentioned in the opinions, presented in Table 1, suggest very different types of trials unfolded in the trafficking and CSA cases. Electronic and physical evidence were mentioned far more often in trafficking than CSA cases, χ²s (1) > 13.01, ps < .001, whereas medical evidence was mentioned more often in CSA cases, χ² (1) = 6.58, p = .01. Trafficking defendants were more likely to have taken the stand, whereas CSA defendants were more likely to have confessed, χ²s (1) ≥ 5.04, ps ≥ .03.

Turning to expert testimony, experts for the prosecution were mentioned significantly more often in trafficking than CSA cases, χ² (1) = 11.40, p < .001. We reviewed the opinions for details on the type of expert and found that trafficking prosecution experts were typically police who described typical controlling behaviors of traffickers (e.g., victims cannot make eye contact with other traffickers, the victim’s lack of “freedom to come and go”). These experts’ comments, per the opinions, described the defendant as a professional pimp operating a business that required control of all aspects of the victims’ lives. CSA prosecution experts, in contrast, tended to be psychologists or social service professionals who discussed delayed disclosure. A much smaller number of opinions mentioned defense experts, which occurred with similar frequency in trafficking and CSA cases. Defense experts in trafficking cases typically focused on the defendants’ personality traits. Some, for example, described the results of personality tests, highlighting how the results revealed the defendant did not possess traits that would lead to the exploitation. Others described common characteristics of traffickers or abusers and explained how the defendant does not possess these characteristics. Of interest, none of the defense experts mentioned in the appellate opinions described errors in memory or suggestibility in either sample.

Finally, we compared the total number of forms of evidence mentioned between case types. Significantly more forms of prosecution-focused evidence were mentioned in trafficking than CSA cases, t (97) = 3.50, p < .001, and more forms of defense-focused evidence were mentioned in CSA than trafficking cases, t (97) = −1.83, p = .04 (Table 1).

Discussion

The goals of the present investigation were to understand characteristics presented in successfully prosecuted criminal cases involving charges of sex trafficking or pandering of a minor, particularly in terms of how the victims are described, and to assess how those characteristics diverge from characteristics presented in successfully prosecuted CSA cases with similarly aged victims. Three key themes emerged, all of which have implications for prosecuting criminal cases involving adolescent victims of trafficking and CSA.

The first theme, which is especially relevant to prosecutors’ decisions, concerns variations in how involved and cooperative the trafficking versus CSAvictims seemed to have been, even in the successful cases we sampled. Some opinions, mostly in the trafficking cases, included explicit comments describing the victims as uncooperative. More often, though, the opinions described behaviors and case characteristics that indirectly suggested the trafficking victims were both uncooperative and cooperative at different points during the case. A near equal number of the CSA opinions made comments about cooperativeness, but the comments largely described the victims as or suggested the victims were cooperative. Several case and victim characteristics are consistent with these rather striking differences in comments about potential uncooperativeness versus cooperativeness between the two types of victims.

One difference between case types that both could be related to cooperativeness and was frequently mentioned in the opinions concerned how the victimization was discovered. The trafficking victims tended to come to the attention of the authorities because they were identified by law enforcement and not on their own accord. When victims do not report abuse themselves, they are often evasive in interviews and may deny abuse even when presented with evidence of its occurrence (Katz, 2013; Nogalska et al., 2021). Both evasiveness and denial could be interpreted as victim uncooperativeness. In contrast, in our sample, the adolescent sexual abuse victims’ disclosures most often initiated the CSA case. Although evidence suggests that adolescent sexual abuse victims often delay disclosing their abuse, once adolescents tell, they tend to maintain their reports (Azzopardi et al., 2019; Goodman-Brown et al., 2003; Malloy et al., 2007). Adolescent sexual abuse victims may weigh the consequences of disclosing carefully and feel ready to maintain their report once they tell, which likely then translates into appearing more cooperative with the authorities.

Compared to the CSA opinions, the trafficking opinions also more often mentioned that victims’ testimony from preliminary hearings or from a hearsay witness was presented in the trial. Prosecutors often express concerns that trafficking victims will fail to appear at key interviews and hearings, for instance, because they feel committed to their trafficker (Farrell et al., 2016). When victims fail to testify at the trial, prosecutors may resort to introducing the victims’ prior statements from the preliminary hearing or from a hearsay witness. Our findings suggest that prior testimony and hearsay were central components of some trafficking criminal trials, as evidenced by the appellate opinions’ inclusion of these case details in the fact summaries. It will be worthwhile to examine systematically in the future whether victim uncooperativeness increases prosecutors’ use of alternative forms of presenting victim statements at trial, especially in trafficking cases, and how such alternative forms of presenting victims’ statements, directly and in conjunction with victim cooperativeness, affect trial outcomes.

With regard to the CSA victims’ cooperativeness, as mentioned, their abuse was most often discovered as a result of their own disclosures, potentially making them more willing to take part in the criminal case. In addition, medical evidence was mentioned in some CSA but not trafficking cases. Having corroborating medical evidence available might have made it easier for victims to maintain their disclosures or fully commit to the case, and medical evidence is consistently linked to prosecutorial success in CSA cases (Brewer et al., 1997). It would be of interest to assess whether medical evidence indeed leads to or is related to victim cooperativeness among adolescent sexual abuse victims. It would also be of interest to explore what types of medical evidence are available and presented in trafficking cases, and whether medical evidence’s inclusion affects trafficking victims’ disclosures or cooperation or the case outcome.

Finally, trafficking and CSA opinions differed substantially in how often caregivers were mentioned. Only five trafficking cases (7%) mentioned caregivers, whereas a much larger percentage of CSA cases did so. With the CSA cases, as well, the appellate opinions described a vast majority of caregivers as supportive. Caregiver support is related to shorter delays between abuse and children’s and adolescents’ disclosures (Koçtürk & Bilginer, 2020), to increases in the likelihood that CSA and adolescent rape cases will be prosecuted (Cross et al., 1994; Duron, 2018; Feeney et al., 2018), and to improved functioning and more positive feelings in CSA victims during and after their case (Goodman et al., 1992). That caregiver support was frequently mentioned in the CSA appellate opinions suggests that caregivers’ presence played a central or important role in the trial decision as well. Whether caregivers are present or supportive of trafficking victims in criminal trials and whether this presence influences the outcome or victims’ cooperativeness are not clear but are important directions for future research on prosecution of trafficking and pandering of minors.

A second theme to emerge across the opinions concerned the characterization of trafficked adolescents as troubled and problematic. Trafficking victims’ histories of delinquent, criminal, and deceptive behaviors were frequently mentioned. At times, character evidence (e.g., the victim lies and is un-trustworthy) might be relevant to the facts of the case. Yet, in many opinions, the links between victims’ delinquent behaviors and charges against the defendant were less clear, a pattern consistent with surveys of prosecutors, who express concerns about irrelevant details about victims’ pasts being admitted into trials (Farrell et al., 2014). Indeed, several opinions mentioned victims’ past runaway behavior, use of illicit drugs, and other types of delinquent behavior, often in combination. One opinion, for example, described a victim’s fights with her mother as being so intense that the mother felt obliged to contact social services. Certainly, many trafficked youth have delinquent histories, with some being identified by law enforcement because of that delinquent behavior (Fedina et al., 2019; Halter, 2010; Hershberger et al., 2018). However, the frequency with which this behavior was mentioned in the trafficking case opinions was striking, particularly when compared to how rarely such behavior was mentioned in the CSA case opinions. Thus, the tendency to label trafficking victims as delinquent youth, which potentially increases the amount of blame placed on them for their victimization, seems pervasive. Not only do laypersons, law enforcement, and court professionals exhibit this tendency (Farrell et al., 2014; Halter, 2010; Winks et al., 2022), but our results suggest that authors of appellate opinions do as well.

Opinions at times, though, also alluded to the victims’ history of exposure to adversity, with this occurring more often with trafficked than sexually abused adolescents. One possibility is that the adversity exposure was presented, at least in part, to mitigate potential harmful effects associated with discussions of a victim’s delinquency or as a reason why the victim had engaged in delinquency. In the trafficking sample, for example, mention of adversity was significantly correlated with the number of delinquent acts mentioned, r (67) = .42, p < .001. Therefore, the trafficking victims seemed to be described in complex ways. Their negative behaviors were featured fairly prominently in trials but so were their histories of prior trauma. How positive and negative depictions of trafficked minors influence jurors’ decisions and how these depictions are offset by other evidence are significant and worthwhile follow-up questions to explore, especially given that our sample included only successful cases.

The third theme concerned the range of manipulative strategies employed by traffickers and CSA perpetrators to induce compliance. Several noteworthy trends emerged. For one, our findings suggest that labeling traffickers as either seduction-focused or intimidation-focused in their manipulation tactics does not adequately capture the varied ways that traffickers manipulate their victims. Some opinions described trafficking defendants as having used violence, coercion, and isolation to manipulate victims. Mentioning of these tactics was expected given prior research that suggests cases in which violence was used to coerce trafficking victims into compliance are more likely to be prosecuted than cases in which no violence was used (Farrell et al., 2014; Feeney et al., 2018). Yet, opinions also frequently mentioned non-intimidation tactics, such as feigned romance, psychological manipulation, and isolation, and no clear pattern emerged regarding combinations of tactics being mentioned, at least at the level of the appellate opinion. In subsequent work, it will be important to ascertain how combinations of manipulation tactics in addition to individual tactics relate to case outcomes.

Next, with one exception, the types of manipulation tactics employed by the trafficking and CSA defendants in our sample were quite similar. This may suggest that adults who exploit youth employ manipulation tactics that target developmental vulnerabilities rather than crime-specific tactics, an issue worthy of further investigation. Of note, a small set of CSA opinions (but no trafficking opinions) mentioned perpetrators enticing victims by exposing them to pornography. Whether this particular manipulation tactic is a strategy employed by a subset of CSA perpetrators who target adolescent victims or is primarily reflective of perpetrators who have been found guilty at trial is also an important direction for future research.

Limitations

Although our research provides valuable new knowledge about criminal cases involving adolescent trafficking and sexual abuse victims, it is not without limitations. First, our sample, which was comprised of successful trial convictions that were appealed, is not representative of all trafficking or CSA criminal cases, especially cases that were never presented to the prosecution and cases presented but then rejected by prosecutors. It is reasonable to assume that cases in which victims were particularly uncooperative are unlikely to be prosecuted, unless there is some clear form of corroborative evidence (see Farrell et al., 2016, for a discussion of this issue). However, our sample is likely representative of most cases that resulted in guilty verdicts in California for two reasons. One is that the state allows an appeal as of right, and the other is that all appeal decisions in the state, not just those certified for publication (i.e., opinions believed to be addressing an important legal issue in need of settling; CA Rule of Court 8.1005), are available in written form to the public. Nontheless, our sampling procedure likely inflated the frequency of victim cooperativeness and the existence of corroborative evidence relative to what one would observe in non-prosecuted or even unsuccessfully prosecuted cases.

Second, legal opinions do not report on all evidence in a case. Instead, the evidence discussed reflects the case aspects that judges saw as most relevant to the case outcome, reason for appeal, and appellate decision. Thus, our findings say about as much about appellate judges’ responses to trafficking and CSA cases involving adolescent victims as our findings do about what occurred at trial. Future research should examine actual trial transcripts, document all forms of evidence, and include acquittals to understand more fully the myriad of ways that victims’ experiences, victims’ cooperativeness, and inclusion of evidence shape how criminal trials involving trafficking and CSA unfold. Finally, because many of our cases involved multiple victims, and victims were clustered within cases, some of our analyses at the victim level violated assumptions of independence. For example, defendants in a single case may have been more likely to treat victims in that case similarly. We would note, however, that we observed substantial variability within cases in the characteristics of different victims, and patterns of findings remained similar when we reconducted analyses randomly selecting only one victim or victim-defendant pair from a given case.

Conclusion

Even with some limitations, our results highlight the complexity of criminal cases involving sexual victimization of minors, especially those who have been subjected to trafficking or pandering. Cases often include multiple victims spanning wide age ranges and, at times, several defendants. Prosecutors, therefore, need to have considerable knowledge concerning trafficking manipulation, perceptions of victim compliance, and perhaps how best to combat stereotypes and misconceptions regarding why trafficked youth behave the way they do. We suspect that, in our sample of successfully prosecuted cases, prosecutors possessed some such knowledge, given that several of the trafficking cases in our sample included potentially uncooperative or inconsistently cooperative victims, and victims who seemed to feel attached to their trafficker. Yet, our cases also often had sufficient physical evidence available and prosecution experts who could explain victim characteristics and victim-trafficker dynamics. Overall, our findings highlight that, with a clear understanding of victim history and victim-perpetrator relationship, it is possible to be successful when prosecuting trafficking cases. Our findings also highlight, though, that an understanding of prosecution of cases involving trafficking of minors is separate from an understanding of prosecution of CSA cases with adolescent victims. Striking differences in the characteristics of victims, defendants, and the cases themselves reveal the need for specialized and unique training for prosecutors working on the two types of cases.