Verified Human Trafficking Allegations Among Single and Dual System-Involved Children: Predicting Initial and Repeat Victimization

Abstract

There is a dearth of research examining repeat human trafficking victimization among children involved with the child welfare system (i.e., single system involvement) and children involved with both the child welfare and juvenile justice systems (i.e., dual system involvement). This study uses longitudinal statewide linked administrative data from the Florida Department of Children and Families (DCF) and Florida Department of Juvenile Justice (DJJ) to investigate both initial and subsequent verified human trafficking allegations among single and dual system-involved children. We conducted logistic regression models to identify youth characteristics, prior DCF experiences, and prior DJJ experiences that predict initial and subsequent trafficking victimization. We also conducted survival analysis to identify time until human trafficking victimization and revictimization. Prior maltreatment, placement history, missing child events, referral without adjudication, community supervision, and residential facility placement predict initial trafficking victimization. These same indicators predict revictimization, except for prior placement history, residential facility commitment, prior physical abuse and prior sexual abuse. Approximately 1 in 5 child victims experience trafficking revictimization; the median time between initial and subsequent revictimization is about 6 months. This research has implications for policy and practice among system-involved children at greatest risk for human trafficking revictimization.

Keywords

child trafficking human trafficking labor trafficking sex trafficking commercial sexual exploitation of children revictimization repeat victimization subsequent victimization child abuse child welfare child maltreatment juvenile justice children minors youth

Introduction

Although there is robust research on risk factors for and predictors of children’s human trafficking¹ victimization, little research exists on predictors of repeat or subsequent human trafficking victimization. Existing research on risk factors finds that involvement in the juvenile justice and child welfare systems and maltreatment predict initial trafficking victimization (Fedina et al., 2019; Franchino-Olsen, 2021). However, awareness of predictors of repeat trafficking victimization is necessary to inform appropriate responses to initial trafficking allegations and effectively prevent future revictimization. Without proper system responses and interventions that address the underlying causes of initial victimization, children may be susceptible to repeat trafficking victimization and its attendant harms.

To our knowledge, this is the only study to examine repeat human trafficking victimization, specifically repeat victimization among children involved with the child welfare system (i.e., single system involvement) and children involved with both the child welfare and juvenile justices (i.e., dual system involvement). Using statewide administrative data from both Florida’s Department of Children and Families (DCF) and Department of Juvenile Justice (DJJ), we examine predictors of initial and subsequent verified human trafficking allegations (VHTA) among single and dual system-involved children.

System Involvement, Childhood Abuse, and Trafficking

Researchers have consistently found a positive association between involvement with the juvenile justice and child welfare systems and experiencing maltreatment with human trafficking victimization (Franchino-Olsen, 2021). This association is consistent across numerous studies that use a diverse array of data sources; measures of system involvement, maltreatment, and victimization; and analysis methods. Many studies have found that justice system involvement is associated with trafficking victimization as either an adult (Nichols & Cox, 2021) or minor (Franchino-Olsen, 2021), with most studies examining sex trafficking victimization. Measures of justice system involvement include incarceration in a juvenile detention center (Chohaney, 2016; Pullmann et al., 2020) and receipt of citations and bench warrants (Cook et al., 2021). These studies range from surveys of street-based sex workers asked to self-report age of entry into sex work (Chohaney, 2016), analysis of juvenile justice data from participants in a juvenile justice system–based specialty court program (Cook et al., 2021), analysis of state juvenile justice and child welfare system administrative data (Cutbush Starseed et al., 2024; Pullmann et al., 2020), and reviews of medical records for 12–18 year-olds with known involvement with law enforcement (Varma et al., 2015).

Similarly, researchers have also consistently found that child welfare or child protective services involvement predicts trafficking victimization (Franchino-Olsen, 2021). Analyses of child welfare system referrals (Bath et al., 2020), life course interviews with adult victims of sexual exploitation (Coy, 2009), and administrative data from statewide child protective services agencies (Gibbs et al., 2018; Havlicek et al., 2016; Pullmann et al., 2020) found that minors with child welfare contact (e.g., placements) are more likely than those without to experience trafficking victimization, with more frequent placements increasing odds of trafficking.

Related, experiencing child abuse and maltreatment is also consistently associated with trafficking victimization. Many studies have found that trafficking victims experience high rates of maltreatment as children (Fedina et al., 2019; Franchino-Olsen, 2021), and minors who experienced abuse were more likely to experience human trafficking than those who did not experience abuse (Gibbs et al., 2018). This is true for sexual abuse (Pate et al., 2021), physical abuse (De Vries & Goggin, 2020), emotional abuse (Fedina et al., 2019), and multiple forms of abuse and adverse childhood experiences (Naramore et al., 2017).

Human Trafficking Revictimization

Despite the robust research that has consistently found an association between juvenile justice involvement, child welfare involvement, and child maltreatment with both child and adult human trafficking victimization, less empirical research has established how these childhood experiences are associated with repeat or subsequent trafficking victimization. Only a small number of published studies examine trafficking revictimization (N = 4), summarized below.²

The first study, an International Organization on Migration (IOM) report, examined the 79 sex and labor re-trafficking cases among adult and minor females from the IOM victimization database. Victims came from all over the world and were most frequently transported to another country (Jobe, 2015). The reasons for their revictimization were attributed to deportation or return migration to their home countries before they were able to receive adequate services and support. Once back in their home country, there was a lack of support, understanding, and services for these victims, and they often faced additional stigma and bias from family and community members that prevented their recovery.

Another study (Chen et al., 2023) focused on 87 adult sex trafficking victims, 23 of whom had been re-trafficked in New York City; the majority of those 87 victims were non-citizen clients of a social service medical clinic. Using retrospective case file reviews, the authors found that re-trafficked women were more likely to have experienced childhood sexual and physical abuse, unstable legal status, and lived with a non-family member. In their implications section, the authors argued that the revictimization of these individuals may have been caused by experiencing “more severe pre-trafficking vulnerabilities or that those vulnerabilities were not mitigated after escaping trafficking the first time” (Chen et al., 2023, p. 12).

The last quantitative study (Kafafian et al., 2021) is an analysis of re-referral data from a domestic child sex trafficking victim service program. Using data from 413 participants who were mostly referred by the child welfare system and 120 participants who were subsequently re-referred, the authors found that re-referrals were driven by experiencing childhood adversity, such as history of abuse, substance misuse issues, missing from care, homelessness, and mental health issues. Participants with multiple system contacts before their first referral were also more likely to be re-referred to this program. The authors discussed how it was likely participants’ previous vulnerability and experiences of maltreatment, running away and homelessness, and system contact that led to their re-referrals by child welfare agency stakeholders—not necessarily new commercial sexual exploitation victimization.

The qualitative study (Nichols et al., 2022) provides some additional insights for these quantitative patterns and the underlying causes of trafficking revictimization. Interviews with 35 social service providers and criminal justice agency staff who interacted with minors who had experienced commercial sexual exploitation revealed how punishment in the juvenile justice system creates conditions for repeat trafficking because the system is unable to meet the needs of children who have experienced sex trafficking, thus letting both the impacts of their victimization and their underlying vulnerabilities go unaddressed. Both types of respondents lamented the lack of effective interventions, health care, social services, residential facilities, and other programs that can help trafficked youth address their vulnerabilities to and impacts from initial victimization to prevent their repeat victimization.

Current Study

The purpose of this study is to examine predictors of both initial and subsequent VHTA among single and dual system-involved children. Using statewide administrative data from Florida’s DCF and DJJ to examine predictors of initial and subsequent VHTA, we fill an empirical gap in two literatures: predictors of initial trafficking victimization and subsequent trafficking victimization among single and dual system–involved minors.

Methods

Data

The data we used in this study are longitudinal administrative data from Florida’s DCF and DJJ, and the study population for this analysis comprised all children born on or after January 1, 1993, who had at least one DCF-recorded maltreatment allegation from January 1, 1993, through February 29, 2020 (N = 1,611,063). We chose the cut-off date of February 2020 to minimize the impacts of COVID-19 on child welfare or juvenile justice practices. Data received from DCF included demographic information and comprehensive information on all maltreatment allegations, removals, and placements from birth through February 2020. Identifiers in the DCF administrative data were matched to DJJ administrative data to facilitate inclusion of DJJ predictors. To preserve the confidentiality of these children, DCF staff sent personally identifiable data to DJJ, and DJJ staff matched the IDs from each system using information such as name, date of birth, and sex. After matching, data related to DJJ referrals, services, and assessments for the children in the DCF sample were deidentified before being securely transferred to the project team for analyses. All youth included in the analysis were less than 18 years old.

The DCF data included the most serious finding of the DCF investigation for each allegation type. The finding categories were (1) verified, where most of the credible evidence indicated the maltreatment resulted from exploitation, abuse, or neglect; (2) not substantiated, where there was not enough evidence to make a conclusion; and (3) not indicated, where there was no evidence to support the claim.

Overall, out of a total of N = 3,829,772 investigations, there were N = 12,354 investigations that included any human trafficking allegation (either “human trafficking CSEC”, “human trafficking labor” or unspecific “human trafficking”). Some investigations (N = 372) had the same DCF reporting date as another human trafficking investigation due to multiple perpetrators or multiple people making a call to the DCF hotline. We combined any investigations of a human trafficking allegation that occurred on the same DCF date into one investigation, resulting in N = 12,167 investigations. Of these investigations, 24% (N = 2875) were a verified human trafficking allegation, whereas 23% (N = 2815) were not substantiated, and 53% (N = 6477) were not indicated. For this study, only verified human trafficking allegations are treated as victimization (and revictimization) experiences.

Measures

The two human trafficking outcomes for this study were (1) whether a child had a VHTA, and (2) whether there was a subsequent VHTA³ of those with an initial VHTA (N = 2276). Since we would be using events that happened before a VHTA as predictors in logistic regression and survival analysis, we created two reference dates. Then we constructed predictor variables using information about events occurring before each reference date.

Reference Dates

For children who had a VHTA, the reference date was the date of the allegation. For those without a VHTA, the reference date was their 18th birthday or March 1, 2020 (the day after the last day of data in the analysis file), whichever was earlier. For survival analyses, children without a VHTA were “censored” at this reference date, meaning the period of time to look for the event (i.e., a VHTA) was stopped prior to finding the event of interest. We also created a second reference date for children who had experienced an initial VHTA. Analogous to the first reference date, the second reference date was the date of their second VHTA. For those with a first VHTA but without a second VHTA, the second reference date was either their 18th birthday or March 1, 2020, whichever came sooner.

Predictors

Demographic predictors included age at the time of each reference date, sex at birth, and a combined race and ethnicity variable. DCF predictors included individual variables for missing child events, out-of-home placements, and allegations. Specifically, we counted the number of DCF placements and any DCF-reported missing child events that started before each reference date. We also reviewed allegation data to see if any neglect, physical abuse, sexual abuse, psychological⁴ maltreatment allegations were received on or before each reference date. For purposes of this analysis, we considered allegation types received on the same date as the VHTA as occurring prior to it.

DJJ predictors included the number of referrals and a variable created from a combination of information on referrals and DJJ services received before each reference date. Specifically, the variable included separate categories for children with (1) a placement in any residential facility with or without community supervision, (2) any community supervision but no residential facility placement, (3) any referrals but no adjudication, and (4) no DJJ referrals.

We used the same process to identify predictor variables happening before the first and second reference dates. Events could have occurred at any point prior to the second reference date, including if it only occurred prior to the first reference date.

Analyses

We conducted descriptive analyses, logistic regression analyses, and survival analyses. Descriptive analyses included descriptive frequencies, percentages, means, and sums of child demographic characteristics at the time of the first and second reference dates and events occurring prior to each reference date. We conducted logistic regression models to identify youth characteristics, prior DCF experiences, and prior DJJ experiences that predict initial and subsequent trafficking victimization. The predictors were age at the reference date; sex, race, and ethnicity of the child; any DJJ involvement prior to the reference date; any prior neglect; any prior sexual abuse; any prior physical abuse; any prior psychological abuse; number of prior placements; number of prior missing child events; and the number of prior DJJ referrals. While logistic regression allows us to model the probability of a binary outcome, it does not address the time until the event occurs. We therefore fit analogous parametric survival models with the same set of predictors as the logistic regression models to predict the time to an event. The outcome variable in the first survival model was time to first VHTA; we censored children who never experienced a VHTA. The outcome variable in the second survival model was time to second VHTA among children who had at least one VHTA; we censored children who never experienced a second VHTA. Preliminary analyses looked at the distribution of time to allegation and compared models with logistic, Gaussian, Weibull, exponential, lognormal, and loglogistic distributions using Akaike’s Information Criterion (Akaike, 1974). The best fitting distribution was the Weibull distribution, which we ultimately used for the models reported in the results section.

Results

Descriptive Statistics

As shown in Table 1, less than 1% of children in the study had a VHTA (N = 2276). However, of those with a VHTA, 18% (N = 399) had a subsequent VHTA. The median number of days from birth until a child’s first VHTA was 5827 days, or about 16 years (see Table 1). The median number of days until a subsequent VHTA was only 193 days, or about 6 months.

Table 1.

Descriptive Statistics for Youth’s History Prior to First and Second VHTA (or Censoring for Those Without a VHTA).

Characteristic		Total N (%)	Initial VHTA victimization			A second VHTA victimization
Characteristic		Total N (%)	No N (%)	Yes N (%)	p-value	No N (%)	Yes N (%)	p-value
Total children		1,611,063 (100)	1,608,787 (100)	2276 (100)		1877 (100)	399 (100)
Demographics
Year of birth	1993–2006	856,938 (53.2)	854,725 (53.1)	2213 (97.2)	<.0001	1816 (96.8)	397 (99.5)	0.0024
Year of birth	2007–2020	754,125 (46.8)	754,062 (46.9)	63 (2.8)		61 (3.2)	2 (0.5)
Age at first HT/censoring	0–10	556,496 (34.5)	556,441 (34.6)	55 (2.4)	<.0001	53 (2.8)	2 (0.5)	<.0001
	11–12	181,485 (11.3)	181,382 (11.3)	103 (4.5)		87 (4.6)	16 (4.0)
	13–14	175,474 (10.9)	174,958 (10.9)	516 (22.7)		379 (20.2)	137 (34.3)
	15–16	159,307 (9.9)	158,264 (9.8)	1043 (45.8)		840 (44.8)	203 (50.9)
	17	538,301 (33.4)	537,742 (33.4)	559 (24.6)		518 (27.6)	41 (10.3)
Age at second HT/censoring	0–10					27 (1.4)	2 (0.5)	<.0001
	11–12					30 (1.6)	5 (1.3)
	13–14					104 (5.5)	68 (17.0)
	15–16					270 (14.4)	217 (54.4)
	17					1446 (77.0)	107 (26.8)
Sex	Female	789,486 (49.7)	787,548 (49.6)	1938 (85.4)	<.0001	1556 (83.3)	382 (95.7)	<.0001
Sex	Male	799,960 (50.3)	799,630 (50.4)	330 (14.6)		313 (16.7)	17 (4.3)
Race and ethnicity	White, non-Hispanic	511,893 (31.8)	511,287 (31.8)	606 (26.6)	<.0001	521 (27.8)	85 (21.3)	<.0001
	Black, non-Hispanic	343,637 (21.3)	342,873 (21.3)	764 (33.6)		586 (31.2)	178 (44.6)
	Other, non-Hispanic	452,213 (28.1)	451,698 (28.1)	515 (22.6)		445 (23.7)	70 (17.5)
	Hispanic	303,320 (18.8)	302,929 (18.8)	391 (17.2)		325 (17.3)	66 (16.5)
DJJ involvement before HT/Censoring
Prior DJJ involvement	Any residential facility w/or w/o community supervision	13,878 (0.9)	13,790 (0.9)	88 (3.9)	<.0001	159 (8.5)	25 (6.3)	<.0001
	Any community supervision w/o residential facility	34,712 (2.2)	34,443 (2.1)	269 (11.8)		278 (14.8)	89 (22.3)
	Any referral but no adjudication	65,758 (4.1)	65,327 (4.1)	431 (18.9)		299 (15.9)	98 (24.6)
	No DJJ involvement	1,496,715 (92.9)	1,495,227 (92.9)	1488 (65.4)		1141 (60.8)	187 (46.9)
Mean number of prior DJJ offenses		0.3	0.3	1.6	<.0001	2.4	2.7	0.1110
DCF involvement before HT/censoring
Any prior DCF involvement		1,609,610 (99.9)	1,607,680 (99.9)	1930 (84.8)	<.0001	1877 (100.0)	399 (100.0)
Any prior sexual abuse allegations		170,225 (10.6)	169,309 (10.5)	916 (40.2)	<.0001	812 (43.3)	220 (55.1)	<.0001
Any prior physical abuse allegations		537,517 (33.4)	536,225 (33.3)	1292 (56.8)	<.0001	1086 (57.9)	291 (72.9)	<.0001
Any prior neglect allegations		1,107,389 (68.7)	1,105,749 (68.7)	1640 (72.1)	0.0006	1398 (74.5)	341 (85.5)	<.0001
Any prior psychological maltreatment allegations		149,265 (9.3)	148,642 (9.2)	623 (27.4)	<.0001	524 (27.9)	168 (42.1)	<.0001
Mean number of prior placements		0.5	0.5	3.1	<.0001	5.0	7.5	0.0006
Mean number of prior missing child reports		0.1	0.0	2.0	<.0001	3.2	5.6	<.0001
Median number of days until first HT				5827 days
Median number of days between first and second HT							193 days

Most children with a VHTA were female (85%) compared to an equal split between sexes among children without a VHTA. Among those with a subsequent VHTA, 96% were female. Among those without a VHTA, only 7% had any sort of DJJ involvement, whereas over a third (35%) of those with a first VHTA and over half (53%) of those with a subsequent VHTA had some sort of previous DJJ involvement. The mean number of prior DJJ delinquency charges for children without a VHTA was 0.3, whereas children with a VHTA averaged 1.6 DJJ delinquency charge per child. For those with a second VHTA, the mean was 2.7 delinquency charges per child prior to the second VHTA.

The pattern for percent of children experiencing prior abuse is similar to DJJ involvement. The percentages experiencing each prior abuse type (physical abuse, sexual abuse, psychological maltreatment, and neglect) were lowest for those with no VHTA, and highest for those with multiple VHTA. For example, one-third of children without a VHTA had a prior physical abuse allegation, over half (57%) of those with an initial VHTA had a prior physical abuse allegation, and almost three-quarters (73%) of children with a subsequent VHTA had a prior physical abuse allegation.

The mean number of prior DCF placements for those without any VHTAs was 0.5, whereas for those with a VHTA the mean number of prior DCF placements was 3.1 placements prior to their first VHTA and 7.5 placements prior to their second VHTA. Children with no VHTA averaged zero missing child events, whereas those with a VHTA averaged two missing events prior to their first VHTA and 5.6 missing child events prior to a second VHTA.

Logistic Regression Models

Logistic regression model results are presented on the scale of the odds ratio (OR) to give an effect size estimate for each predictor. The results shown in Table 2 are adjusted odds ratios (AOR) from models that included all other variables listed in Table 2.⁵ The likelihood of having a VHTA increased with age, with children aged 15 or 16 years more than 40 times more likely (AOR = 42.34) than children aged 10 or younger to have a VHTA. A smaller AOR at age 17 than the other groups was because these children were at risk for a shorter period; specifically, they were censored at age 18 (i.e., if they were in the sample until they aged out at age 18 and never had a VHTA as a child). Girls were six times (AOR = 6.04) more likely than boys to have a VHTA. Non-Hispanic Black children and non-Hispanic children of other races were around twice as likely (AOR = 1.68 and 2.14, respectively) to have a VHTA, and Hispanic children were slightly more likely (AOR = 1.33) than non-Hispanic White children to have a VHTA. Children experiencing any kind of DJJ involvement were more likely to experience a VHTA than those without, ranging from children who had experienced any prior residential facility placement (AOR = 6.40), to those with a referral but not adjudication (AOR = 6.54), to those with prior community supervision (AOR = 7.76). Children with prior physical (AOR = 1.43), psychological (AOR = 1.67), or sexual (AOR = 2.54) abuse were more likely than those without to have a VHTA. Each DCF-recorded missing child event increased the adjusted odds of having a VHTA by 1.04. For example, child with 10 missing child events was almost twice as likely to have a VHTA (AOR for 10 missing child events = 1.04¹⁰ = 1.48) than children without an event.

Table 2.

Logistic Regression Results.

Characteristic	Initial VHTA (yes vs. no) (N = 1,589,446)			Subsequent VHTA (yes vs. no) (N = 2268)
Characteristic	Adjusted odds ratio	95% CI	p-value	Adjusted odds ratio	95% CI	p-value
Demographic variables
Age at HT/censoring (ref: 0–10 years old)
11–12	5.66	(4.07–7.86)	<.0001	2.60	(0.45–14.93)	0.2850
13–14	24.60	(18.57–32.59)	<.0001	5.02	(1.12–22.41)	0.0348
15–16	42.34	(32.12–55.82)	<.0001	5.89	(1.35–25.73)	0.0184
17	3.56	(2.66 - 4.75)	<.0001	0.49	(0.11–2.16)	0.3476
Sex (ref: Male)
Female	6.04	(5.35–6.82)	<.0001	3.79	(2.20–6.53)	<.0001
Race and ethnicity (ref: White, non-hispanic)
Black, non-Hispanic	1.68	(1.51–1.88)	<.0001	2.19	(1.57–3.05)	<.0001
Other, non-Hispanic	2.14	(1.89–2.42)	<.0001	1.21	(0.81–1.81)	0.3563
Hispanic	1.33	(1.17–1.52)	<.0001	1.51	(1.01–2.26)	0.0463
DJJ variables
Number of prior DJJ referrals	1.03	(1.01 - 1.05)	0.0008	0.99	(0.95–1.04)	0.7117
Prior DJJ involvement (ref: No DJJ involvement)
Any residential facility placement with or without community supervision	6.40	(4.62–8.88)	<.0001	1.19	(0.56–2.53)	0.6524
Any community supervision without residential facility placement	7.76	(6.47–9.32)	<.0001	1.95	(1.22–3.12)	0.0052
Any referral but no adjudication	6.54	(5.77–7.42)	<.0001	1.66	(1.16–2.39)	0.0060
DCF variables
Any prior neglect (ref: no)	0.93	(0.84–1.02)	0.1370	0.98	(0.67–1.42)	0.9029
Any prior physical abuse (ref: no)	1.43	(1.30–1.56)	<.0001	1.17	(0.85–1.60)	0.3389
Any prior sexual abuse (ref: no)	2.54	(2.32–2.78)	<.0001	1.10	(0.84–1.44)	0.4922
Any prior psychological abuse (ref: no)	1.67	(1.50–1.85)	<.0001	1.34	(1.01–1.78)	0.0426
Number of prior DCF placements	1.01	(1.01 - 1.02)	<.0001	1.00	(0.99–1.01)	0.9515
Number of prior missing child events	1.04	(1.04–1.05)	<.0001	1.02	(1.00–1.03)	0.0099

The second logistic regression model included those children who had any VHTA to predict a second VHTA. Children aged 13–16 were more than 5 times as likely (AOR = 5.02 for 13–14 year olds and 5.89 for 15–16 year olds) as 0–10 year olds to have a subsequent VHTA. Girls were still more likely (AOR = 3.79) more likely than boys to have a subsequent VHTA. Non-Hispanic Black children and Hispanic children were more likely (AOR = 2.19 and AOR = 1.51, respectively) than non-Hispanic White children to have a subsequent VHTA. Both children with prior DJJ community supervision and those with a referral but no adjudication were about twice as likely to have a subsequent allegation as children with no prior involvement (AOR = 1.95 and AOR = 1.66, respectively). Children with a residential facility placement did not have significantly different adjusted odds of having a subsequent VHTA than children with no prior DJJ involvement. The only prior abuse type that was significant was psychological abuse where children who had any prior psychological abuse were slightly more likely (AOR = 1.34) than those without to have a second VHTA. Similar to predicting the first VHTA, each DCF-recorded missing child event increased the adjusted odds of having a VHTA by 1.02.

Survival Models

Table 3 shows the results of the survival analysis looking at time to the first VHTA, and time from the first to the second VHTA. Results of the survival models are shown by hazard ratio (HR). Like the OR, an HR greater than 1 indicates a predictor is a risk factor for a VHTA and an HR less than 1 indicates a predictor is a protective factor. However, they are interpreted differently from the OR. For example, an HR = 2 for a binary predictor indicates that children with that predictor have VHTA at twice the rate of children without the presence of that binary predictor. The results shown in Table 3 are adjusted hazard ratios (AHR) from models that included the other variables listed in Table 3.⁶

Table 3.

Parametric Survival Analysis Results.

Characteristic	Time to initial VHTA (N = 1,589,446)			Time to subsequent VHTA (N = 2268)
Characteristic	Adjusted hazard ratio	95% CI	p-value	Adjusted hazard ratio	95% CI	p-value
Demographic variables
Age at HT/censoring (ref: 0–10 years old)
11–12	4.53	(3.26–6.30)	0.0000	3.12	(0.60–16.18)	0.1761
13–14	18.97	(14.32–25.12)	0.0000	8.71	(2.11–35.91)	0.0028
15–16	30.98	(23.50–40.84)	0.0000	7.21	(1.77–29.30)	0.0059
17	3.30	(2.47–4.40)	0.0000	0.83	(0.20–3.39)	0.7938
Sex (ref: Male)
Female	6.03	(5.34 – 6.80)	0.0000	3.90	(2.37 – 6.44)	0.0000
Race (ref: White, non-hispanic)
Black, non-Hispanic	1.63	(1.46–1.82)	0.0000	1.72	(1.32–2.24)	0.0001
Other, non-Hispanic	2.91	(2.58–3.29)	0.0000	1.25	(0.90–1.75)	0.1872
Hispanic	1.43	(1.26–1.63)	0.0000	1.44	(1.04–1.99)	0.0287
DJJ variables
Number of prior DJJ referrals	1.03	(1.01–1.04)	0.0023	0.99	(0.95–1.03)	0.5831
Prior DJJ involvement (ref: No DJJ involvement)
Any residential facility placement with or without community supervision	0.17	(0.12–0.23)	0.0000	0.96	(0.51–1.82)	0.9004
Any community supervision without residential facility placement	1.17	(0.90–1.53)	0.2400	1.60	(0.97–2.62)	0.0649
Any referral but no adjudication	0.97	(0.72–1.30)	0.8298	1.55	(0.86–2.78)	0.1430
Any prior neglect (ref: no)	0.76	(0.69 - 0.84)	0.0000	0.97	(0.71–1.34)	0.8642
Any prior physical abuse (ref: no)	1.34	(1.23–1.47)	0.0000	1.22	(0.94–1.58)	0.1381
Any prior sexual abuse (ref: no)	2.37	(2.17–2.60)	0.0000	0.94	(0.76–1.17)	0.5856
Any prior psychological abuse (ref: no)	1.62	(1.46–1.79)	0.0000	1.20	(0.96–1.50)	0.1097
Number of prior DCF placements	1.01	(1.01–1.01)	0.0001	1.00	(0.99 – 1.00)	0.3331
Number of prior missing child events	1.03	(1.03–1.04)	0.0000	1.01	(1.00–1.02)	0.0065

In looking at time to the first VHTA, the trafficking rate for children aged 13–14 is 19 times the trafficking rate of children aged 0–10 (AHR = 18.97), whereas the rate of 15- and 16-year-olds is over 30 times those of children aged 10 or younger (AHR = 30.98). For the first VHTA, the trafficking rate for girls is more than 6 times the trafficking rate of boys (AHR = 6.03). Children with any prior sexual abuse, physical abuse, or psychological abuse allegations experience a VHTA quicker than children who did not have these abuse allegations (AHR = 2.37 for sexual, AHR = 1.34 for physical and 1.62 for psychological). However, children with a prior neglect allegation have an AHR less than one (AHR = 0.76), meaning those without a prior neglect allegation have a higher rate of VHTA.

In the analysis of children with at least one VHTA, the AHR for the second VHTA were similar. The AHR for children aged 13–14 years is over 8 times the rate of children aged 10 or younger (AHR = 8.71), and the rate for children aged 15–16 years is over 7 times the rate (AHR = 7.21) of children aged 10 or younger. Non-Hispanic Black children (AHR = 1.72) and Hispanic children (AHR = 1.44) had rates of VHTA higher than Non-Hispanic White children.

Discussion

Using longitudinal, linked, administrative data from the DCF and DJJ, we conducted logistic regression models to identify predictors of initial and subsequent trafficking victimization and survival analysis to identify time until VHTA among system-involved youth. Logistic regression models predicted whether the child had a VHTA and, of those with an verified allegation, whether a second verified allegation occurred. Survival models predicted (1) time to first VHTA and (2) time to second VHTA among children who had at least one VHTA.

Study findings show that within this sample there were N = 2276 children with a VHTA. Of those with a verified allegation, 18% (N = 399) had a subsequent VHTA. This staggering rate of revictimization is alarming and signifies the need for sustained, intensive prevention strategies that are designed to 1) mitigate the contextual circumstances that places the child at risk for trafficking; and 2) decrease the child’s risk factors and increase the child’s protective factors. The median number of days from birth until a child’s first VHTA was 5827 days, or 16 years. The median number of days until a subsequent VHTA was only 193 days, or about 6 months; the high prevalence of revictimization, coupled with the rapid onset of revictimization, suggests the need for expedited child welfare investigations occurring simultaneous to evidence based, comprehensive service provision.

Demographic Characteristics

Although one literature review of domestic minor sex trafficking studies suggests that demographic factors do not function as significant predictors of minor sex trafficking (Choi, 2015), our study findings indicate that age, sex, race, and ethnicity all significantly predict initial and subsequent trafficking victimization.

Findings from this study of system-involved children suggest that as age increases, so does the likelihood of human trafficking victimization. Children aged 15 or 16 years are 42 times more likely than children aged 10 or younger to be victimized and at a rate over 30 times that of children 10 or younger. Although commensurate with other study results (Havlicek et al., 2016; Varma et al., 2015), this finding should be interpreted with caution given the shortage of validated human trafficking screening and identification tools for young children. For example, it is possible that the likelihood of being trafficked is similar across age groups but that human trafficking service providers do not yet have valid tools and interview techniques that are developmentally appropriate for screening and identifying young children (Greenbaum, 2017).

Findings from other studies consistently identify girls as being at higher risk of trafficking victimization (Naramore et al., 2017; Polaris, 2018), and this study is no exception. Compared with boys, girls were 6 times more likely to have an initial VHTA and about 4 times (OR = 3.79) more likely to have a subsequent allegation.

Although consistent with other literature, this finding should once again be interpreted with caution: First, it is likely that the number of male victims is widely underestimated (Greenbaum, 2017; Nichols et al., 2022) owing to a number of policy and practice reasons, such as less frequent screening and identification for boys (ECPAT USA, 2013). Second, child labor trafficking research is in its infancy, but early research suggests within group differences between child sex trafficking and child labor trafficking, namely that labor trafficked children, juxtaposed with sex trafficked children, were more likely to be male than female (Aba, 2019; Greenbaum et al., 2022). Because this study’s sample is disproportionately comprised of sex trafficking cases, gender estimates may be biased.

Race and ethnicity emerged as significant predictors of both initial and subsequent trafficking victimization, with Black children, other children of color, and Hispanic children experiencing increased risk for initial trafficking victimization—and Black children and Hispanic children (but not other children of color) experiencing increased risk for subsequent trafficking victimization—when compared with White, non-Hispanic children. Countless studies document being Black, a child of color, or Hispanic as increasing risk for trafficking victimization for both system-involved and non–system-involved youth).

Although the overrepresentation of Black children and children of color in the current and other system-involved samples may impact findings, the data are concerning nonetheless - and highlight the nationwide need for all youth serving agencies, institutions, and organization to apply an anti-racist lens (Cook et al., 2022; Klein & Merritt, 2014; Nichols et al., 2023) to anti-trafficking policy and practice, both by interrogating and addressing (1) bias on the part of those who refer or report and investigate the maltreatment (see “Bias Model” [Harris, 2021]) and (2) increased exposure to familial and community risk, such as economic vulnerability, housing instability, residential crowding, community violence, and child care burden (see “Risk Model” [ Feely & Bosk, 2021]). Still, the universality of potential trafficking victimization should be noted; trafficking is not endemic to any one community.

Prior Maltreatment

Results from the current study suggest that children with prior sexual abuse, prior psychological abuse, and prior physical abuse, (but not prior neglect), compared with those without, were more likely to have experienced an initial VHTA. However, only prior psychological abuse longitudinally predicted trafficking revictimization. Prior sexual, psychological, and physical abuse (in that order) significantly hastened the onset of initial trafficking victimization; however, surprisingly, prior neglect functioned as a protective factor, delaying the onset of initial trafficking victimization.

Numerous studies have reported high levels of child neglect (Hopper, 2017; Naramore et al., 2017; Pate et al., 2021; Ulloa et al., 2016), physical abuse (De Vries & Goggin, 2020; Hopper et al., 2017), psychological abuse (Fedina et al., 2019), and sexual abuse (De Vries & Goggin, 2020; Hopper, 2017; Naramore et al., 2017; Pate et al., 2021; Reid et al., 2017) among trafficked children, although studies may vary on the type or number of maltreatment types that most strongly predict trafficking (Naramore et al., 2017; Ulloa et al., 2016). Surprisingly, though, findings from this study suggest that although prior abuse (sexual, psychological, and physical) function as a significant predictor of child trafficking victimization, neglect does not. One possible explanation for this difference involves trauma bonding, defined as an emotional attachment between an abuser and a victim (Reid et al., 2013). It is possible that child victims of active sexual, psychological, and physical abuse experience trauma bonds with their abuser, reinforced through cycles of violence that involve alternating positive and negative interactions, thereby conditioning the victim to experience gratitude for positive interactions and self-blamed for negative interactions, or even internalize the perpetrator’s viewpoint (Casassa et al., 2022). Child victims of abuse may normalize this harmful dynamic and therefore be especially vulnerable to trafficking victimization (Sanchez et al., 2019). Juxtaposed to “active” forms of abuse characterized by trauma bonding, it is possible that child victims of neglect may not experience alternating positive and negative interactions, experiencing a consistent absence of interactions instead, and therefore may not experience trauma bonding or may experience it to lesser degrees. Future research should examine the role of trauma bonding among all types of child maltreatment (e.g., as a correlate, predictor, mediator, and moderator).

Interestingly, although sexual abuse was the strongest predictor of initial victimization, psychological maltreatment was the only significant child maltreatment predictor of subsequent trafficking victimization. One systematic review (Vrubliauskaitė, 2018) of psychological concerns as risk factors for child sex trafficking suggest the following: that low self-esteem, denigration of self and others, dysfunctional attachment models, learned maladaptive problem solving strategies, and post-traumatic stress disorder – which may emerge after traumatic experiences such as child sexual abuse, multiple placements in foster care, and related missing child events, may converge to increase risk for trafficking victimization. It is therefore plausible that said psychological concerns (i.e., which increased vulnerability for initial trafficking victimization) may be exacerbated by the primary trafficking victimization experience, thereby increasing a child’s risk for subsequent victimization. Future research should investigate the role of psychological maltreatment, and psychological outcomes germane to other types of child abuse and neglect, as risk factors for primary and repeat trafficking victimization experiences.

Finally, vexing as it may be, models suggest that all types of active abuse (i.e., sexual, psychological, physical) hasten the onset of subsequent victimization, but that neglect functions as a protective factor, delaying the onset of repeat victimization. This counter-intuitive finding should be replicated through other studies and interrogated through future research prior to informing policy or practice or being cited.

History of Child Placements

Any, but especially more frequent, out of home placements are positively associated with an exponentially increased risk profile for VHTAs. First, the average number of prior DCF placements for those without any VHTA was 0.5, whereas for those with an allegation the average number of prior DCF placements was 3.1 placements prior to their first VHTA. Moreover, children with only 1 VHTA had an average of 5 placements, whereas those with multiple VHTA averaged 7.5 placements prior to their second allegation. Further, prior child placements emerged as a significant predictor of initial VHTA. However, prior child placements fell out of the model as a predictor of subsequent victimization, suggesting its predictive power is attenuated in cases of revictimization. Previous studies point to child placements (e.g., family placements, congregate care placements) as a risk factor of human trafficking (Havlicek et al., 2016; Pullmann et al., 2020), and the current study punctuates that risk. This study’s finding underscore the need for child welfare systems to improve child placement stability through evidence-based policy and programmatic strategies (Fischer et al., 2013), such as leveraging predictive modeling involving placement matching (Trudeau et al., 2023), increasing training and support for caregivers (McDonald et al., 2023), and providing early and intensive support to wrap the child in a more stable and supportive environment (McDonald et al., 2023).

Missing Child Events

Our data bear a similar pattern for missing child events as for prior maltreatment and prior child placements: Children with more than one VHTA had more missing child events than those with one VHTA, and those with one VHTA had more missing child events than those with no VHTA. Each missing child event increased the odds of an initial VHTA occurring. For example, a child with 10 missing child events is almost twice as likely to have a VHTA. Missing child events are well established in the literature as a strong risk factor for trafficking victimization (Latzman et al., 2019). Our current study now also documents missing child events as not only a longitudinal predictor for initial trafficking, but also establishes missing child events as a longitudinal predictor of trafficking revictimization. Florida DCF notes that more than 95% of missing episodes, regardless of age, are caused by a youth running away from a placement (Florida Department of Children and Families et al., 2008).⁷ These findings point to the urgent need for implementation of 1) effective strategies designed to prevent children from running (Chor et al., 2002); and 2) universal screening for both sex and labor trafficking after recovering children missing from care (Gambon et al., 2020).

Child Welfare Summary

In sum, prior maltreatment (i.e., sexual abuse, psychological abuse, and physical abuse), placement history, and missing child events function as significant longitudinal predictors of initial trafficking victimization, whereas only prior psychological maltreatment and missing child events predict repeat trafficking victimization among system-involved youth. Importantly, among children with repeat trafficking victimization allegations, children who had experienced any number of prior missing child events—compared with those who had not—had significantly shorter windows of time between their initial and subsequent trafficking victimization. Surprisingly, having experienced any prior maltreatment or any prior DCF placements did not significantly shorten the time between initial and repeat trafficking victimization. These collective findings emphasize the need to escalate targeted secondary and tertiary trafficking intervention activities among children with initial trafficking victimization experiences, especially children who also have a history of any number of prior missing child events.

Juvenile Justice Involvement

This study documents DJJ involvement as a significant predictor of both initial and subsequent child trafficking victimization among child welfare involved youth. The strongest DJJ predictors for initial trafficking victimization within this sample included (1) prior community supervision, (2) referral without adjudication, and (3) prior residential facility placement, in that order. The strongest longitudinal DJJ risk factors for subsequent trafficking victimization within this sample included (1) prior community supervision and (2) referral without adjudication; prior residential facility placement was not a significant longitudinal predictor of trafficking revictimization. There are several possible explanations for this series of findings: First, children who are under community supervision are being monitored or watched more closely and/or more frequently; their probation officers may therefore be more attentive to risks for or early detection of impending trafficking victimization. Second, children who are referred to the delinquency system but whose cases are pending action or do not involve any sort of judicial action may not be monitored closely or at all after their juvenile justice arrests, which would make subsequent victimization less likely to come to the attention of either system. Third, children who are in residential facilities are incarcerated, thereby diminishing opportunities for revictimization. This latter finding, however, should be interpreted with extreme caution: studies have repeatedly documented that arresting or incarcerating youth to “protect them” from further trafficking victimization causes egregious harm, further traumatizing the youth and penalizing them with an “offender” status and criminal record (Zabresky, 2013) To date, Safe Harbor laws, now enacted by 30 states (Shared Hope International Institute for Justice & Advocacy, 2023), including Florida (Subramanian, 2020), are designed to redirect child victims of trafficking from the criminal justice system and into the child welfare system for victim-centered, trauma-informed, age-appropriate, and evidence-based wrap-around services.

Juvenile Justice Summary

This study fills an important empirical gap by documenting that children with prior DJJ involvement (defined here as any community supervision without residential facility placement or any referral but no adjudication) were about twice as likely to have a subsequent allegation than children with no prior involvement. Although prior research has investigated human trafficking victimization among justice-involved youth, most relied on small or selective samples, such as children from specific residential facilities (O'Brien et al., 2017) or intervention court (Cook et al., 2021), or they relied on arrest records for commercial sex involvement as their measure of human trafficking (Naramore et al., 2017). To date, only one study involving a small sample (N = 83) of dual system–involved children who have experienced human trafficking exists, documenting a relationship between a detention episode and human trafficking victimization (Pullmann et al., 2020). These findings underscore the pivotal opportunity Departments of Juvenile Justice have in preventing both initial and/or subsequent child trafficking victimization: in-depth trafficking victimization trainings for juvenile justice staff; clear policy and practice involving screening high-risk youth; and careful monitoring of justice-involved youth may improve primary, secondary, and tertiary prevention of trafficking victimization.

Repeat Trafficking Victimization

Study findings indicate that approximately 1 in 5 system-involved children experienced trafficking revictimization. This staggering rate of revictimization is alarming and signals the need for sustained, intensive secondary and tertiary prevention for children placed at risk for human trafficking victimization. Specifically, programs and service models should implement data driven strategies designed to mitigate the family, community, social, and ecological circumstances the child is placed in De Vries et al. (2024).]. In addition, programs and service models should target a reduction in the child’s modifiable risk factors (Franchino-Olsen, 2021) and an increase in the child’s modifiable protective factors.

Limitations

The primary limitation facing this study involves the confined number of demographic, social, child welfare, and juvenile justice indicators available within the linked administrative dataset we leveraged for analyses. Race was confined to “White,” “Black,” or “Other,” which impedes our ability to understand variability involving racial subgroups. Moreover, data on gender identity and sexual orientation were unavailable. In addition, no data on nationality or primary language spoken were available. Although these demographic limitations are not necessarily unique to our study, state agencies should consider transcending conventional norms involving demographic data collection. In addition, within the linked administrative data, there was a minimal number of relevant DCF and DJJ indicators available beyond prior abuse, placements, and missing child reports as well as referrals, community supervision, and residential facility placement. Furthermore, there are limitations because of the identification of the specific kind of trafficking experienced. The small number of labor trafficking cases further limits this study. Finally, because these data are specific to youth within Florida, findings from this population may not be generalizable to child welfare and juvenile justice settings in other states. Limitations notwithstanding, this study contributes to the scant body of research involving initial and repeat human trafficking victimization among system-involved children.

Conclusions

This study reports child characteristics and system experiences that increase risk for an initial and subsequent VHTA, thereby equipping policy-makers and practitioners to (1) more effectively prevent initial trafficking victimization among children under system supervision; (2) more effectively prevent subsequent trafficking victimization by expediting both initial trafficking victimization investigation and immediate service provision, particularly among those at even greater risk; and (3) provide comprehensive services designed to heal and protect children from future harm. Additional research is needed to better understand both predictors of and time between initial and repeat trafficking victimization experiences among children who are system-involved. Future research should place greater attention on collecting more nuanced data pertaining to race, gender identity, sexual orientation, nationality, primary language spoke, system experiences, and seek to include an equal distribution of sex and labor trafficking cases.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute of Justice (Award Number, 2018-MU-MU-0033).

ORCID iD

Marianne Kluckman

Notes

References

Aba

L. M.

(2019). Defining Child Trafficking for Labor Exploitation, Forced Child Labor, and Child Labor.

Akaike

(1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19(6), 716–723. https://doi.org/10.1109/tac.1974.1100705

Bath

E. P.

Godoy

S. M.

Morris

T. C.

Hammond

Mondal

Goitom

Farabee

Barnert

E. S.

(2020). A specialty court for U.S. youth impacted by commercial sexual exploitation. Child Abuse and Neglect, 100, Article 104041. https://doi.org/10.1016/j.chiabu.2019.104041

Casassa

Knight

Mengo

(2022). Trauma bonding perspectives from service providers and survivors of sex trafficking: A scoping review. Trauma, Violence, and Abuse, 23(3), 969–984. https://doi.org/10.1177/1524838020985542

Chen

V. H.

Beauchemin

E. L.

Cuan

I. T.

Sadana

Olulola-Charles

Leschi

J. E.

Ades

(2023). Sex trafficking in New York City and vulnerabilities to re-trafficking. Journal of Interpersonal Violence, 38(21-22), 11501–11519. https://doi.org/10.1177/08862605231183452

Chohaney

M. L.

(2016). Minor and adult domestic sex trafficking risk factors in Ohio. Journal of the Society for Social Work and Research, 7(1), 117–141. https://doi.org/10.1086/685108

Choi

K. R.

(2015). Risk factors for domestic minor sex trafficking in the United States: A literature review. Journal of Forensic Nursing, 11(2), 66–76. https://doi.org/10.1097/JFN.0000000000000072

Chor

K. H. B.

Luo

Dworsky

Raman

Courtney

M. E.

Epstein

R. A.

(2002). Development and validation of a predictive risk model for runaway among youth in child welfare. Children and Youth Services Review, 143, Article 106689. https://doi.org/10.1016/j.childyouth.2022.106689

Cook

M. C.

P. D.

Garcia

J. J.

(2022). Addressing racism in the domestic minor sex trafficking of Black girls: The role of public health critical race praxis. Public Health Reports, 137(1_suppl), 10S–16S. https://doi.org/10.1177/00333549211058735

10.

Cook

M. C.

Talbert

R. D.

Thomas

(2021). A longitudinal study of justice characteristics among girls participating in a sex trafficking court program. Health Justice, 9(1), 1. https://doi.org/10.1186/s40352-020-00127-1

11.

Coy

(2009). Moved around like bags of rubbish nobody wants’: How multiple placement moves can make young women vulnerable to sexual exploitation. Child Abuse Review, 18(4), 254–266. https://doi.org/10.1002/car.1064

12.

Cutbush Starseed

Stephen

Marianne

Lilly

Samuel

(2024). Child Trafficking as a Predictor of Subsequent Juvenile Justice Involvement. Journal of Aggression, Maltreatment & Trauma. https://doi.org/10.1080/10926771.2024.2433679

13.

De Vries

Baglivio

Reid

J. A.

(2024). Examining individual and contextual correlates of victimization for juvenile human trafficking in Florida. Journal of Interpersonal Violence, 39(23-24), 4748–4771. https://doi.org/10.1177/08862605241243332

14.

De Vries

Farrell

(2018). Labor trafficking victimizations: Repeat victimization and polyvictimization. Psychology of Violence, 8(5), 630–638. https://doi.org/10.1037/vio0000149

15.

De Vries

Goggin

K. E.

(2020). The impact of childhood abuse on the commercial sexual exploitation of youth: A systematic review and meta-analysis. Trauma, Violence, and Abuse, 21(5), 886–903. https://doi.org/10.1177/1524838018801332

16.

ECPAT USA . (2013). And Boys Too: An ECPAT-USA discussion paper about the lack of recognition of the commercial sexual exploitation of boys in the United States. https://d1qkyo3pi1c9bxcloudfrontnet/00028B1B-B0DB-4FCD-A991-219527535DAB/1b1293ef-1524-4f2c-b148-91db11379d11pdf

17.

Fedina

Williamson

Perdue

(2019). Risk factors for domestic child sex trafficking in the United States. Journal of Interpersonal Violence, 34(13), 2653–2673. https://doi.org/10.1177/0886260516662306

18.

Feely

Bosk

E. A.

(2021). That which is essential has been made invisible: The Need to bring a structural risk perspective to reduce racial disproportionality in child welfare. Race and Social Problems, 13(1), 49–62. https://doi.org/10.1007/s12552-021-09313-8

19.

Fisher

P. A.

Mannering

A. M.

Van Scoyoc

Graham

A. M.

(2013). A translational neuroscience perspective on the importance of reducing placement instability among foster children. Child Welfare, 92(5), 9–36.

20.

Florida Department of Children and Families, Community Based Care Providers, & Florida Department of Law Enforcement/Missing Endangered Persons Information Clearinghouse . (2008). Missing children guide: Location, stabilization, and prevention.

21.

Franchino-Olsen

(2021). Vulnerabilities relevant for commercial sexual exploitation of children/domestic minor sex trafficking: A systematic review of risk factors. Trauma, Violence, and Abuse, 22(1), 99–111. https://doi.org/10.1177/1524838018821956

22.

Gambon

T. B.

Gewirtz O'Brien

J. R.

Committee on Psychosocial Aspects of Child and Family Health, Council on Community Pediatrics . (2020). Runaway youth: Caring for the nation's largest segment of missing children. Pediatrics, 145(2), Article e20193752. https://doi.org/10.1542/peds.2019-3752

23.

Gibbs

D. A.

Henninger

A. M.

Tueller

S. J.

Kluckman

M. N.

(2018). Human trafficking and the child welfare population in Florida. Children and Youth Services Review, 88, 1–10. https://doi.org/10.1016/j.childyouth.2018.02.045

24.

Greenbaum

Sprang

Recknor

Harper

N. S.

Titchen

(2022). Labor trafficking of children and youth in the United States: A scoping review. Child Abuse and Neglect, 131, Article 105694. https://doi.org/10.1016/j.chiabu.2022.105694

25.

Greenbaum

V. J.

(2017). Child sex trafficking in the United States: Challenges for the healthcare provider. PLoS Medicine, 14(11), Article e1002439. https://doi.org/10.1371/journal.pmed.1002439

26.

Harris

M. S.

(2021). Racial bias as an explanatory factor for racial disproportionality and disparities in child welfare. In Dettlaff

A. J.

(Ed.), Racial disproportionality and disparities in the child welfare system Child Maltreatment (11, pp. 141–158). Springer. https://doi.org/10.1007/978-3-030-54314-3_8

27.

Havlicek

Huston

Boughton

Zhang

(2016). Human trafficking of children in Illinois: Prevalence and characteristics. Children and Youth Services Review, 69, 127–135. https://doi.org/10.1016/j.childyouth.2016.08.010

28.

Hopper

E. K.

(2017). Polyvictimization and developmental trauma adaptations in sex trafficked youth. Journal of Child and Adolescent Trauma, 10(2), 161–173. https://doi.org/10.1007/s40653-016-0114-z

29.

Jobe

(2015). The causes and consequences of re-trafficking: Evidence from the IOM human trafficking database. https://publications.iom.int/books/causes-and-consequences-re-trafficking-evidence-iom-human-trafficking-database

30.

Kafafian

De Vries

Farrell

Goldfarb

Bouchard

(2021). Understanding factors associated with re-referral of youth for commercial sexual exploitation. Child Abuse and Neglect, 117, Article 105092. https://doi.org/10.1016/j.chiabu.2021.105092

31.

Klein

Merritt

D. H.

(2014). Neighborhood racial & ethnic diversity as a predictor of child welfare system involvement. Children and Youth Services Review, 41, 95–105. https://doi.org/10.1016/j.childyouth.2014.03.009

32.

Latzman

N. E.

Gibbs

D. A.

Feinberg

Kluckman

M. N.

Aboul-Hosn

(2019). Human trafficking victimization among youth who run away from foster care. Children and Youth Services Review, 98, 113–124. https://doi.org/10.1016/j.childyouth.2018.12.022

33.

McDonald

K. P.

Fisher

Connolly

(2023). Building a specialized model of care for youth involved in sex trafficking in child welfare: A systematic review and interviews with experts-by-experience. Child abuse and Neglect, 135, Article 105987. https://doi.org/10.1016/j.chiabu.2022.105987

34.

Naramore

Bright

M. A.

Epps

Hardt

N. S.

(2017). Youth arrested for trading sex have the highest rates of childhood adversity: A statewide study of juvenile offenders. Sexual Abuse, 29(4), 396–410. https://doi.org/10.1177/1079063215603064

35.

Nichols

A. J.

Cox

(2021). A pilot study comparing sex trafficking indicators exhibited by adult and minor service populations. Journal of Human Trafficking, 9(2), 194–211. https://doi.org/10.1080/23322705.2021.1898173

36.

Nichols

A. J.

Preble

Cox

(2022). A state-level analysis of demographic characteristics and sex trafficking experiences of survivors. Journal of Human Trafficking, 10(1), 51–68. https://doi.org/10.1080/23322705.2021.2016268

37.

Nichols

A. J.

Preble

Cox

Nichols

Amann

(2023). An examination of sex trafficking by race and ethnicity in the state of Missouri on state and local levels: Implications for anti-oppressive practice. Journal of Human Trafficking, 19(6), 1–16. https://doi.org/10.1080/23322705.2023.2225993

38.

O’Brien

J. E.

Givens

Leibowitz

G. S.

(2017). Domestic minor sex trafficking among adjudicated male youth: Prevalence and links to treatment. Children and Youth Services Review, 82, 392–399. https://doi.org/10.1016/j.childyouth.2017.09.026

39.

Pate

S. S.

Anderson

V. R.

Kulig

T. C.

Wilkes

Sullivan

C. J.

(2021). Learning from child welfare case narratives: A directed content analysis of indicators for human trafficking. Children and Youth Services Review, 121, Article 105838. https://doi.org/10.1016/j.childyouth.2020.105838

40.

Polaris . (2018). Hotline statistics 2017. Retrieved October 6 from. https://polarisproject.org/2017statistics

41.

Pullmann

M. D.

Roberts

Parker

E. M.

Mangiaracina

K. J.

Briner

Silverman

Becker

J. R.

(2020). Residential instability, running away, and juvenile detention characterizes commercially sexually exploited youth involved in Washington State's child welfare system. Child Abuse and Neglect, 102, Article 104423. https://doi.org/10.1016/j.chiabu.2020.104423

42.

Reid

J. A.

Baglivio

M. T.

Piquero

A. R.

Greenwald

M. A.

Epps

(2017). Human trafficking of minors and childhood adversity in Florida. American Journal of Public Health, 107(2), 306–311. https://doi.org/10.2105/AJPH.2016.303564

43.

Reid

J. A.

Haskell

R. A.

Dillahunt-Aspillaga

Thor

J. A.

(2013). Contemporary review of empirical and clinical studies of trauma bonding in violent or exploitative relationships. International Journal of Psychology Research, 8(1), 37–73.

44.

Sanchez

R. V.

Speck

P. M.

Patrician

P. A.

(2019). A concept analysis of trauma coercive bonding in the commercial sexual exploitation of children. Journal of Pediatric Nursing, 46, 48–54. https://doi.org/10.1016/j.pedn.2019.02.030

45.

Shared Hope International Institute for Justice & Advocacy . (2023). Safe Harbor laws: Legal protections for trafficked children & youth. https://reportcards.sharedhope.org/wp-content/uploads/2023/09/Safe-Harbor-Fact-Sheet.pdf

46.

Subramanian

(2020). Beyond detention-as-protection for child sex trafficking victims. Berkeley. Journal Of Gender, Law and Justice, 35, 137–167.

47.

Trudeau

K. J.

Yang

Kraus

D. R.

(2023). Predicting successful placements for youth in child welfare with machine learning. Children and Youth Services Review, 153, Article 107117. https://doi.org/10.1016/j.childyouth.2023.107117

48.

Ulloa

Salazar

Monjaras

(2016). Prevalence and correlates of sex exchange among a nationally representative sample of adolescents and young adults. Journal of Child Sexual Abuse, 25(5), 524–537. https://doi.org/10.1080/10538712.2016.1167802

49.

Varma

Gillespie

McCracken

Greenbaum

V. J.

(2015). Characteristics of child commercial sexual exploitation and sex trafficking victims presenting for medical care in the United States. Child Abuse and Neglect, 44, 98–105. https://doi.org/10.1016/j.chiabu.2015.04.004

50.

Vrubliauskaitė

(2018). Psychological factors increasing the risk for adolescent girls to become victims of human trafficking: A systematic review. Socialinis Darbas, 16(1), 129–138.

51.

Zabresky

(2013). Creating a safe harbor for Florida’s children: An overview of Florida’s legislative evolution in domestic minor sex trafficking. Florida State University Law Review, 40, 415–440.