Evidence-Based Human Trafficking Policy: Opportunities to Invest in Trauma-Informed Strategies

Abstract

Human trafficking is a critical social issue characterized by chronic trauma among victims, and frequently preceded by traumatic experiences that contribute to risk of victimization. Therefore, the research-based practice of trauma-informed care is a highly appropriate lens for both prevention and intervention. This work examines federal legislation in the United States related to human trafficking for references to trauma, as well as how the use of research could implicitly direct public policy responses toward trauma-informed approaches. Legislation on human trafficking has risen substantially since 1989, and the use of research and trauma language within these policies has also observed substantial increases. While the use of trauma language was associated with limited progression in the policy process, legislation using research language was more likely to pass out of Committee and become enacted. Moreover, legislation may leverage research in ways that have the potential to bolster trauma-informed practice among human trafficking victims. Specifically, research can be used to describe the problem and causal mechanisms (e.g., impact of trauma), guide “best practice” for service delivery, and generate knowledge through studies and evaluations that guide future policy. Therefore, human trafficking legislation that implicitly guides trauma-informed practice via the use of research may be particularly promising for the field.

Introduction

Survivors of human trafficking—who have been exploited for labor or services through the use of force, fraud, or coercion—experience trauma at particularly high rates (Gozdziak & Collett, 2005). Victimization involves substantial recurring trauma, and many survivors have extensive histories of trauma prior to becoming trafficked (Naramore, Bright, Epps, & Hardt, 2017; Reid, 2012; Zimmerman, Hossain, & Watts, 2011). Leading theory and practice have begun to recognize the utility of trauma-informed and trauma-responsive practice for this population (e.g., Hopper, 2017), augmenting the dominant “three P’s” framework (i.e., protection, prosecution, prevention; Lagon, 2015; Zimmerman et al., 2011). However, the extent to which U.S. federal policy efforts have been guided by related research or practice remains unclear. Although prior research has investigated the explicit use of trauma language in U.S. legislation (Purtle & Lewis, 2017), bills may also be written in ways that draw upon research findings or processes that guide effective practice. Therefore, exploring how research is used in legislation pertaining to a specific group of trauma victims—human trafficking survivors—has the potential to shed light on ways that public policy could advance trauma-informed approaches.

This work considers how public policies pertaining to human trafficking could leverage trauma-informed approaches, both explicitly (e.g., specific mentions) and implicitly (i.e., via the use of research evidence). This work describes the applicability of a trauma-informed lens in the context of human trafficking policy and the potential for the use of research in legislation to indirectly impact trauma-informed practices. We then describe the current work, which (a) reviews U.S. federal legislation for key language suggesting consideration of trauma-informed and evidence-based perspectives, (b) examines the inclusion of these perspectives in the policymaking process over time, and (c) explores how policies might be constructed to reinforce trauma-responsive practices or effective practice overall.

Trauma and Human Trafficking

Trauma experienced by trafficked persons is associated with a range of deleterious outcomes (e.g., Jonson-Reid, Kohl, & Drake, 2012; Mersky, Topitzes, & Reynolds, 2012; Sousa et al., 2011), including heightened vulnerability for future victimization, which is consistent with the perpetual cycle of traumatic experiences among human trafficking victims (Hopper, 2017; Naramore et al., 2017; Reid, 2012; Zimmerman et al., 2011). Many victims experience ongoing trauma and interpersonal abuse, including psychological abuse, physical and sexual violence, and deprivation of basic needs (Hopper, 2017; U.S. Department of State, 2000). Such long-term trauma contributes to an exceedingly high rate of post-traumatic stress disorder among victims (Greenbaum, 2014; Hopper, 2017). Moreover, many victims have extensive histories of adverse experiences and service system utilization (e.g., foster care) prior to victimization (Hopper, 2017; Naramore et al., 2017; Reid, 2012; Zimmerman et al., 2011). While there is some concern that a focus on trauma may lead to negative, deficit-based assumptions of victims (Tseris, 2019), existing literature suggests that a trauma perspective is highly appropriate for both preventing and responding to human trafficking victimization (Hemmings et al., 2016).

Increasingly, research has identified strategies for working with victims in a manner that can avoid further trauma and prevent downstream negative outcomes (Ahn et al., 2013; Hopper, 2017). The literature base on human trafficking is relatively nascent, although the number of publications has grown substantially since 2000 with a particular increase in empirical work after 2010 (Russell, 2018). While studies primarily describe risks and outcomes associated with human trafficking, little research has investigated the effectiveness of aftercare services for victims (Gozdziak & Collett, 2005; Heffernan & Blythe, 2014; Zimmerman et al., 2011). Although limited research means that the effectiveness of trauma-informed practice with human trafficking survivors remains unclear, a relatively robust literature that led to the development of a trauma-informed perspective may be conceptually considered as a best practice for various survivors of trauma. This perspective recognizes the pervasive impact of trauma among many of those interacting with human service systems, as well as the risk of unintentionally retraumatizing or deterring treatment among trauma victims (Harris & Fallot, 2001; Huang et al., 2014). More recently, an outgrowth of this work has begun to recognize trauma-responsive perspectives, which consider systemic changes within organizations or policies to meet the needs of trauma victims. Despite clear implications for policy development, the uptake on a trauma-responsive framework has been minimal in federal legislation, whereas trauma-informed language is occasionally used. Specifically, the Trafficking Victims Housing Act of 2017 (S.2305) is the first and only bill to date that mentions trauma-responsive approaches in U.S. federal legislation. Therefore, consistent with current policy frameworks, we consider trauma-informed approaches alongside those that are trauma-responsive. Furthermore, understanding how legislation might leverage trauma perspectives may support the evaluation of related policies in the future—a noteworthy objective given that the effectiveness or impact of trauma-informed or trauma-responsive policy is not yet known.

Trauma-informed approaches involve processes for recognizing the signs of trauma and designing service systems to accommodate the special needs of trauma survivors (Harris & Fallot, 2001; Huang et al., 2014). Service systems and organizations adopt trauma-informed approaches by emphasizing awareness of trauma symptoms, victims’ unique interests and strengths, emotional safety, freedom of choice, hope, resources, trustworthiness, and transparency (Bowen & Murshid, 2016; Heffernan & Blythe, 2014; Hopper, 2017). Recent efforts have aimed to apply trauma-informed approaches through services for human trafficking victims, including therapy (e.g., Hopper, 2017) and case management (e.g., Heffernan & Blythe, 2014). However, most health workers are not trained to work specifically with this population (Zimmerman et al., 2011), and few survivors seek out, receive, or continue services (Haynes, 2006; Sheldon-Sherman, 2012) which is consistent with the challenges of serving trauma survivors (Harris & Fallot, 2001).

Difficulty in addressing trauma may be particularly relevant to trafficked persons, as a substantial number of victims may receive no services, corresponding with the historical prosecutorial emphasis involving identifying victims by law enforcement (Haynes, 2006). This policy framework expects victims to prove severe occurrence of being trafficked to qualify for services and cooperate with law enforcement on investigation despite justifiable fear that traffickers may retaliate or that the survivor may be prosecuted for crimes committed while being trafficked (e.g., illegal immigration, prostitution; Haynes, 2006). In fact, threats imposed by standard criminal justice practice have the potential to re-traumatize individuals with significant histories of trauma (SAMHSA, 2014). In general, trafficked persons encounter myriad touchpoints in community settings (e.g., health care, social services), which heightens the need for interorganizational and agency partnerships. Such partnerships are most effective when there is mutual and genuine intent across multiple stakeholders to support transformative impact, account for market conditions (i.e., availability of trafficked persons and demand among perpetrators) and use data to assess the problem and monitor progress (Lagon, 2015). Moreover, comprehensive, coordinated primary prevention efforts should involve local, data-driven collaborations that that base strategy on measureable community risks and protective factors (Long et al., 2018). Therefore, efforts to address human trafficking are necessarily complex and multifaceted; however, the use of research evidence in policies pertaining to trafficked persons is also expected to correspond with trauma-responsive practice, which would move service delivery and policies toward victim-centered and trauma-informed approaches, such as those that address barriers to service access (e.g., deterrence of treatment among trauma victims; Substance Abuse and Mental Health Services Administration, 2014).

Despite a growing focus on “trauma-informed” and “trauma-responsive” approaches, the namesake does little to guide research-informed practice, as definitions and quality of implementation vary widely (Becker-Blease, 2017). While some have called for better integration of trauma-informed approaches in public policy, there is little guidance on how to do so (Bloom, 2016; Bowen & Murshid, 2016; Purtle & Lewis, 2017). One proposed model for achieving a victim-centered approach to trafficking within existing legislative frameworks is to build collaborative partnerships between community stakeholders (e.g., law enforcement, service providers, survivors) to improve the identification of victims, protect victims’ rights, and support their rehabilitation, as well as train stakeholders on theoretical best practices for combatting human trafficking (Sheldon-Sherman, 2012). Such efforts present an opportunity to build community capacity for implementing trauma-informed approaches guided by research-based technical support.

Expanding Effective Practice via Public Policy

In the United States, an iterative public policy process involves agenda setting (i.e., defining policy problems and priorities), designing and debating policy alternatives (e.g., legislation), and then interpreting and implementing newly enacted policies through the executive branch (Kingdon & Thurber, 1984; Sabatier & Weible, 2014). Each of these phases could have implications for infusing research-based concepts in policy. For instance, references to trauma-informed practice in U.S. legislation have increased since 2009, when the Domestic Minor Sex Trafficking Deterrence and Victims Support Act of 2010 (S.2925) was introduced (Purtle & Lewis, 2017). Bills may reference trauma symbolically by noting the importance of the issue, or they may also leverage trauma in more material ways by allocating resources or altering processes (Purtle & Lewis, 2017).

Even legislation that does not explicitly reference trauma could have implications for trauma-informed practice if the language used in bills requires administrative agencies to interpret provisions that require supported programs to use effective or research-based practice. In fact, some argue that administrative agencies are better positioned than Congress to allocate resources corresponding with an evolving evidence basis (Haskins & Margolis, 2015), suggesting that laws requiring interpretation of current research, broadly, could indirectly support trauma-informed and other effective practices. Fields with emerging evidence, like human trafficking intervention research, may require flexibility in interpreting the evolving evidence basis to allow federal efforts to be directed in ways corresponding with the “best available” evidence until the field grows capacity for rigorous scientific trials (Head, 2016). Legislation could, for instance, encourage production and review of comparison-group studies, which would provide guidance on the level of early evidence expected.

Although interpreting evidence is essential to generating knowledge that shapes policy decisions, research findings are rarely self-evident. Knowledge is defined by the dominant values of those who have the power and privilege of engaging in decision-making processes (Nutley, Walter, & Davies, 2007; Rallis & Rossman, 2012). However, in the absence of imperfect research-based information, policy decisions will nevertheless be made on the basis of varied inputs from stakeholders (e.g., special interest groups; Kingdon & Thurber, 1984). Research use is not a panacea for achieving social impact, but it may reflect an advancement in attempts to improve effective governance.Dominant values for using rigorous research evidence to guide decision-making have shepherded a growing bi-partisan emphasis on evidence-based policy (Evidence-based Policy Commission; Haskins & Margolis, 2015) and correspond with the emerging field investigating how policymakers use research evidence (e.g., Nutley et al., 2007).

Key examples of “evidence-based” legislation illustrate how reviews of existing research can be used to guide standards for determining service effectiveness and eligibility for grant programs (e.g., Maternal Infant Early Childhood Home Visiting Program; Haskins & Margolis, 2015). However, little work has systematically investigated a range of approaches for writing research evidence or methodologies into legislative language. Further, while significant descriptive research has examined policymakers’ motivations for using research throughout the policy process (Nutley et al., 2007), mechanisms for leveraging research in legislation remain unclear. Specifically, research may be used instrumentally to directly inform policy development or conceptually to indirectly influence how issues are understood. Alternatively, tactical use applies research findings to justify or oppose existing policy proposals. Policies may also impose use through requirements or expectations for research use or involve process use by monitoring, evaluating, or studying issues to inform future policy development (Nutley et al., 2007).

Current Study

This work emerged from the recognition that legislation may leverage research evidence or processes that have indirect implications for trauma survivors, including the way victimization is understood and addressed. We sought to expand on prior investigations that examined trauma-informed practice in U.S. legislation (Purtle & Lewis, 2017) and how policymakers use research evidence (Nutley et al., 2007) by exploring the potential for a trauma-informed perspective to be advanced implicitly in legislation via the use of research evidence. Human trafficking policy was of particular interest because it was anticipated that research-based provisions would have a high likelihood of aligning with a trauma-based perspective. For this work, we define research broadly, and its use included provisions that directly or indirectly leveraged scientific knowledge or concepts, or systematic, empirical, and evaluative methods and processes. This study was guided by three research questions:

1. How has the use of research and trauma language changed over time in human trafficking legislation?

2. Is human trafficking legislation more likely to progress in the legislative process when it references trauma or research evidence?

3. In what ways has human trafficking legislation used research findings or processes?

Method

The analyses for this work occurred in multiple stages, beginning with a keyword analysis followed by content analysis of bills randomly sampled and stratified to represent different types of keywords. A data collection tool (Quorum, 2018) with completely searchable, full-text copies of bills and resolutions was used to identify federal legislation (n = 1,056) introduced between the 101st and the 114th Congresses (1989–2016) using Boolean searches for keywords and phrases relevant to both human trafficking and research use. Primary keywords, provided in Table 1, were then used to capture variations in specific phrases. Research keywords were expected to capture implicit use of trauma-informed approaches and resulted in sampling more legislation than those that explicitly referenced the phrase “trauma-informed” (n = 16) or trauma in general (n = 78; including derivatives of trauma, such as traumatic and trauma-informed). Table 2 provides frequencies of keyword types found within sampled legislation, as well as the mean and standard deviation for the sum of different keywords found within the bill.¹

Table 1.

Primary keywords used to identify legislation

Research use
Research evidence	Adjectives	Verbs	Human trafficking
Evidence	-Informed	Demonstrate(s)	Trafficking or smuggling
Research	-Based or based on	Suggest(s)	-in persons, humans, or children
Studies or study	Experimental	Found or has/have found	Sex or labor
Scientific(ally)	Peer-reviewed	Show or shows	Modern slavery
Data	Rigorous		NOT drug trafficking
Empirical	Randomized

Table 2.

Frequencies and mean use of keywords

	N (%)	M (SD)
Use of keywords
Trauma	78 (7.4)
Trauma-informed	16 (1.5)
Any research keywords	455 (43.1)	0.99 (1.64)
Analysis	38 (3.6)
Methods	134 (12.7)
Research	89 (8.4)
Findings	79 (7.5)
Direct	267 (25.3)
Bill status
Out of committee	307 (29.1)
Enacted	110 (10.4)

Note: n = 1,056; any research keywords = bills that used at least one keyword phrase, and a sum of different keyword phrases that were used in the bill (this does not reflect a total sum of keywords when the same keyword was used more than once within legislation).

Research Keyword Analysis

Analysis began by assessing how the use of research keywords in human trafficking legislation has changed over time and whether the use of such keywords predicted a bill’s progress in the legislative process. Initially, 206 keywords or Boolean phrases were used to identify legislation that referred to studies, research, science, and evidence-base, as well as scientific methods. To refine the selection of keywords, removing those that were either seldom used or overused (i.e., outliers), the mean use of each keyword was computed as the sum of legislation using the keyword out of the sum of all legislation. Fifty-two keywords were retained because their use fell within one standard deviation of the mean. These remaining research keywords were grouped into one of five conceptual categories (each piece of legislation may contain multiple keywords across categories):

1. Analysis keywords referred to statistical analyses including correlations or significant findings (e.g., significantly correlated, comparative analysis).

2. Method keywords referred to the process of obtaining, collecting, or managing data or conducting studies (e.g., data system, statistical data).

3. General Research keywords directly referred to different types of studies or research without necessarily referencing findings (e.g., academic research, latest research).

4. Findings keywords referred to what research or studies have suggested or concluded or references to scientific literature (e.g., peer-reviewed, results of the study).

5. Direct keywords included explicit references to “evidence-based”, “best practice”, and other phrases that did not fit well in the other categories.

Trends in the use of trauma and research keywords were examined over time and used to predict incremental progress of the bill in the legislative process using logistic regression. Specifically, whether the bill passed out of Committee was of particular interest since most legislation “dies” in Committee (Shapiro, 2009). Similarly, the likelihood of enactment was examined as a dependent variable. These analyses controlled for whether the legislation was a resolution, which are often nonbinding and used to recognize issues, people, or events; resolutions were more likely to be passed out of Committee (χ² = 13.91, p < .001) and become enacted than bills (χ² = 111.44, p < .001). Further post hoc analysis examined passage out of Committee and enactment in relation to different types of research keywords used in each bill.

Content Analysis

The content analysis aimed to explore how research was used in legislative text. Drawing from the bills identified in the keyword analysis (i.e., excluding resolutions), a sample of bills was selected at random and stratified to improve representation of the five keyword categories. Moreover, the most recent version of the bill was used for analysis. The final sample is listed in Supplemental Materials. Because bills can be exceedingly lengthy, only sections related to research use and human trafficking were coded and analyzed (Grimmer & Stewart, 2013). Two coders were trained to identify both direct (e.g., using evidence keywords) and indirect references (e.g., concepts such as trauma or risk) to research use and extracted provisions from the sampled bills that indicated research use in relation to human trafficking policy. For instance, indirect research use considered language characterizing known risk or protective factors associated with human trafficking (Long et al., 2018). Subsequently, inductive analyses employed an open-coding scheme to generate thematic categories (Elo & Kyngäs, 2008) regarding how human trafficking legislation has used research processes or findings. After independent thematic analysis, coders met to discuss findings, resolve discrepancies, and achieve consensus on thematic results.

Results

Keyword Analysis

Over one thousand pieces of legislation (n = 1,056) were identified as containing human trafficking keywords—the first two were introduced in the 102nd Congress (1991–1992; 0.02% of all legislation) and the proportion of human trafficking legislation increased to 2.01% in the 114th Congress (2015–2016). Among human trafficking legislation, the use of trauma or research keywords increased over time (Fig. 1). In this time period, references to trauma or trauma-informed keywords within human trafficking legislation increased from 0% to 9.9% and 4.1%, respectively. In contrast, the proportional use of research keywords increased from 0% to 47.7% between the 102nd and 114th Congress. The rise in using trauma keywords began in the 107th Congress (2001–2002), whereas the rise in research keywords began earlier and increased the most between the 102nd and 105th Congresses (1991–1998). Subsequent to those initial increases, the percentage increase in trauma and trauma-informed keywords was relatively stable, whereas the percentage of legislation using research keywords varied substantially and peaked during the 110th and 112th Congresses (2007–2012; 52.3% and 52.9%, respectively).

Fig. 1.

Trends in the use of keywords pertaining to research and trauma. Note: The x-axis indicates the Congressional session during which bills were introduced. The y-axis indicates the total number of human trafficking bills using specified keywords.

Research keyword categories were further examined for trends (Fig. 2) and indicated that Direct keywords (e.g., best practice, evidence-based) were used most frequently and Analysis keywords were least prevalent. However, the proportional use of each research keyword category varied tremendously between Congresses. For instance, General Research keywords were used in no legislation in the 106th Congress (1999–2000), contrasting with use in 23.1% of legislation in the 105th Congress (1997–1998).

Fig. 2.

Trends in the use of research keyword categories. Note: The x-axis indicates the Congressional session during which bills were introduced. The y-axis indicates the total number of human trafficking bills using specified keywords.

Two logistic regressions were used to assess whether the likelihood of the legislation’s passage out of Committee and enactment into law was associated with the use of the word trauma or the sum of research keywords. Legislation using trauma keywords was 48% less likely to pass out of Committee than other legislation when accounting for the legislation type as a resolution and the sum of research keywords used (Table 3). Each time a research keyword was used in a piece of legislation, it was 38% more likely to pass out of Committee and 21% more likely to become enacted. Post hoc analyses of different research keyword categories revealed that legislation using General Research keywords (χ² = 7.37, p = .01) and Direct keywords (e.g., best practice; χ² = 35.81, p < .001) was more likely to pass out of Committee than legislation not using those types of keywords.

Table 3.

Likelihood of legislation progression based on keywords

	Out of committee		Enactment
	OR	p	OR	p
Step 1
Intercept	0.37	<.001	0.07	<.001
Resolution	1.85	<.001	7.34	<.001
Trauma	0.8	.42	0.56	.35
Step 2
Intercept	0.26	<.001	0.05	<.001
Resolution	2.55	<.001	9.12	<.001
Trauma	0.52	.03	0.44	.18
Sum of evidence keywords	1.38	<.001	1.21	.002

Note: n = 1,056; sum of evidence keywords = the sum of different keyword phrases that were used in the bill (this does not reflect a total sum of keywords when the same keyword was used more than once within legislation). OR = Odds Ratio.

Content Analysis

Detailed results from the content analysis, including quotes from sampled bills, are provided in Supplemental Materials. Key findings involved the use of research evidence in human trafficking legislation for three primary purposes: (a) problem definition, (b) problem solution development, and (c) research and evaluation.

Problem Definition

Research and data were used to describe the problem, including the magnitude of the problem and associated consequences (e.g., psychological trauma, HR.2522), as well as risk factors and causal pathways that have implications for public policy (e.g., poverty, HR.4114; foster care, HR.4058). Sometimes, these descriptions used tactical language indicating urgency (e.g., “heinous activity,” S.553) and appeared to educate legislators in order to justify legislative action. Oftentimes, research or data were indirectly referenced to describe the phenomenon without identifying a source. Less frequently, legislation specifically mentioned studies or experts as a source of information.

Problem Solution Development

Both existing research on specific types of programs and conceptual frameworks drawn from research evidence were directly applied to the development, selection, and implementation of service programs. Provisions sometimes required evidence reviews that identified effective programs, governmental agencies, and strategies, which were then used to guide investment in specific strategies and procedures based on evidence of effectiveness. Some grant funding was tied specifically to expectations that the funds be used for purposes with an evidence basis and those that could demonstrate a measurable impact (e.g., “to establish baseline data and monitor progress,” HR.4114, in achieving verifiable goals through implementing “innovative strategies,” S.553). The phrase “best practice” was used to describe practices deemed effective that should be prioritized for government support. Tactical language was also used in research-oriented provisions to encourage immediate action for addressing human trafficking in ways proposed by the bill.

Research and Program Evaluation

Measuring, tracking, and assessing outcomes were a widespread aim found in the sampled human trafficking legislation. Specifically, legislation was used to create knowledge and data repositories about human trafficking and related interventions, necessitate measurable indicators for evaluation, require reports on process and outcomes, call for assessments of cost-effectiveness, and generally expect the use of best practices for program evaluation. Specific studies or methodologies (e.g., “an update of the 2001 Port of Entry Infrastructure Assessment Study,” S.1033) were expected to create scientifically sound information, and sometimes those were to be funded independent of specific programs or services, whereas other times were used to require evaluation of government-funded programs (e.g., “the impact of State safe harbor laws and associated services on the re-victimization of victims of trafficking…the recovery of victims, victim outcomes, and prosecutions of traffickers,” HR.2473). Sometimes, evaluation methods were vaguely specified, whereas other evaluation requirements specified indicators for tracking outcomes (e.g., “determine the baseline prevalence in target populations and outcomes in order to periodically assess progress in reducing prevalence,” S.553; “the annual number and percentage of Federal cases related to human trafficking,” HR.2473), service quality (e.g., “measurements that prioritize the quality of services,” HR.4114), and/or program capacity, implementation, and utilization (e.g., resource and staffing needs, HR.830; participation in victim services, HR.2473). Furthermore, program monitoring and reporting requirements expected grantees to follow through on using or disseminating the information gleaned through evaluation (e.g., by reporting results and recommendations to Congress, HR.4058). Finally, terms such as cost-effectiveness and best practice were used to further indicate expectations that grantees use valid scientific methods to inform the most practical and successful strategies for addressing human trafficking through future legislative actions, research, and services to victims.

Use of Trauma Keywords

Four of the sampled bills used trauma language, and two of those passed out of Committee (Table S1). In one of these bills, “psychological trauma” was used in a conceptual description of the effects of “modern-day slavery” (HR.2522) in order to define the problem. Two of these bills proposed “trauma-informed” services (HR.2473; S.2646), drawing on research in a conceptual way to develop a problem solution. One of those bills also aimed to develop knowledge about serving human trafficking victims (HR.2473). Finally, though one bill used the word trauma to propose increasing the number of trauma centers along the U.S.-Mexico border (S.1033), this was not coded as research use because no research-based or conceptual claim connected access to emergency medical care with the intent to address human trafficking.

Discussion

The findings from this study offer valuable insights regarding how trauma-informed practice could be advanced through legislation. The existing and emergent research basis pertaining to trauma-informed practice creates an opportunity because research-based provisions can indirectly support practices that measurably demonstrate effectiveness. This lens is particularly salient because, despite the proliferation of trauma-informed practice, the quality of implementation may vary considerably and stray from research-based practices with the most potential for effectiveness (Becker-Blease, 2017). The growth of related research (Russell, 2018) may further support such opportunities and correspond with the emergent emphasis of using research evidence in public policy (Commission on Evidence-based Policymaking, 2017). The current findings reveal that human trafficking legislation using research may be more likely to progress in the policy process, particularly when referencing the scale-up of “best practice.” This work further illustrates how such references may be incorporated into legislation that aims to serve human trafficking survivors, a segment of trauma victims for which the evidence-basis is relatively nascent.

This work identified opportunities for framing trauma-responsive policy in ways that may increase the likelihood of advancing in the policy process, specifically, by making reference to research findings or processes. Although human trafficking legislation was less likely to pass out of Committee when using trauma keywords, legislation using research keywords was more likely to advance. However, the negative trend for trauma legislation was only significant when accounting for the use of research keywords, suggesting that legislation using trauma language but not using research keywords may face particular difficulty passing out of Committee. Therefore, emphasizing the research basis behind the trauma-informed movement may strengthen the credibility of relevant policy efforts. However, it is unclear why trauma language fared poorly in the legislative process. Aside from a discouraging interpretation that legislators may be deterred from policies that emphasize trauma, attributes of proposed laws that are not measured here may explain the limited progress of human trafficking legislation using trauma language. For instance, such legislation may be prone to propose increases in discretionary funding compared to other types of human trafficking bills, which would make trauma bills more expensive and less viable for enactment. Moreover, results should be interpreted cautiously as the low frequency of trauma keyword usage may make these analyses susceptible to spurious effects.

Opportunities for framing trauma-informed approaches around “best practice” were apparent in both quantitative and qualitative findings. Specifically, the Direct keywords largely pertained to disseminating, scaling up, or implementing effective practices, and legislation containing this language made greater progress in the legislative process. The in-depth analysis of a selection of bills revealed that best practice may refer to both service delivery and standards for program evaluation. Moreover, both evidence review and evaluation processes often determined what was deemed effective and eligible for programmatic grant funding. Therefore, trauma-informed services may qualify for such grant programs if efficacy has been demonstrated in empirical literature or effectiveness can be measurably demonstrated in practice. Although the bills that were reviewed did not contain “tiered evidence” provisions that often require rigorous empirical studies to demonstrate efficacy as a condition for service funding eligibility (e.g., randomized controlled trials; Haskins & Margolis, 2015), the field of trauma-informed research should anticipate a growing demand for such intervention research among specific populations (e.g., human trafficking victims).

Another theme throughout this work pertains to General Research keywords, including direct references to studies or research (e.g., academic research; recent studies). Legislation using these keywords was more likely to move out of Committee, and these words were used for several different purposes that map onto the policy process: problem definition, policy design, and program review (Sabatier & Weible, 2014). Such keywords may be used to legitimize the description of the problem, authorize study funding, justify general action or a specific policy solution, or develop evidence review procedures. The diverse pathways in which empirical studies were leveraged, despite limited human trafficking research, illuminate numerous opportunities for drawing upon trauma research in public policy. For instance, reference to descriptive research may frame policy solutions geared toward trauma-informed care, or evidence review processes that determine standards of care could adapt to a growing and evolving evidence basis pertaining to trauma-informed interventions.

Another salient theme identified was the tendency for bills to rely on research processes to cultivate knowledge about best practices and use that information in forming a response to the problem. The emphasis on developing a research basis and evaluating current efforts is consistent with the challenges of addressing a social issue for which the empirical basis is extremely limited (Davy, 2016; Sheldon-Sherman, 2012). For instance, current data sources for prevalence, service utilization, and outcomes are flawed, which limits our knowledge on how to best mitigate human trafficking (e.g., Gozdziak & Collett, 2005; Konrad, Trapp, Palmbach, & Blom, 2017; Russell, 2018). Therefore, explicit demands for specific data and measurable outcome indicators are consistent with the priority to develop data and evaluation infrastructure (Tyldum, 2010) corresponding with the intent to determine policy effectiveness and promote accountability over how taxpayer dollars are spent (Head, 2016). Such administrative research infrastructure is intended to fuel future policy initiatives (Sheldon-Sherman, 2012) and could have the potential to inform the delivery of trauma-informed care via service system touch points (e.g., health care; child welfare).

Study findings also align with prior work investigating how policymakers use research evidence (Nutley et al., 2007). For instance, research on trauma may be used conceptually in legislation to justify various policy actions. Alternatively, legislation may directly use research to inform problem solutions, such as trauma-informed services, which is consistent with instrumental research use. In contrast, trauma-informed approaches may be implicitly supported by problem solving approaches that are less prescriptive and impose research use by requiring review processes or evaluations to demonstrate service effectiveness. Moreover, although it is not possible to determine whether research is used to drive a preexisting solution in the context of written legislation, tactical language (e.g., heinous) hints at political use of research evidence (Nutley et al., 2007), and such assertions reinforce the notion of the conceptualization of human trafficking as a traumatic experience.

Limitations

This work builds upon prior work that investigated the explicit use of trauma language in federal U.S. legislation by exploring how research language might implicitly support trauma-informed care within legislation targeting a specific group of trauma survivors—human trafficking victims. Although the focus on human trafficking enabled the examination of implicit instances of trauma-informed policy, this narrow focus limits the generalizability of the work. Specifically, other policy areas may vary in the current knowledge base, existence of an established policy framework, or encounter more ideological disputes (Head, 2016). Furthermore, little attention has been given to the prevention of human trafficking; therefore, the majority of bills sampled likely address the issue after its occurrence—this is more akin to trauma-specific than trauma-informed policymaking (Bloom, 2016). Additionally, though study results are specific to U.S. federal policy and are not generalizable to state legislatures or international policy, lessons learned may be of value in other legislative contexts as legislative language is often drawn from state policies and applied in other states, at the federal level, or beyond (Jansa, Hansen, & Gray, 2015). Such replication of policy efforts also occurs across multiple Congresses (e.g., reintroducing the same bill in a new session) and via companion bills in both chambers, which contributes to unavoidable redundancy between legislation sampled for the current study.

The keyword analysis is marked by strengths and limitations. The current work provides some evidence of convergent validity of the selected keywords by documenting associations with enactment and an upward trend in research words that corresponds with a growing emphasis on evidence-based policy. Additionally, distinct associations with research keywords compared to trauma language offer discriminant validity of the keywords used. However, keywords may have pulled legislation marginally related to human trafficking or may not have identified all instances of research use. The keywords used in analyses are limited by challenges of sensitivity and specificity and, in particular, are not likely to detect indirect references to research in which studies or sources are not mentioned specifically. Such indirect references to research were coded in the content analysis; however, it was beyond the scope of this paper to check the veracity of such claims to ensure the merit of the research basis or original source. Moreover, this work should largely be deemed as descriptive and illustrative, but limited in its implications for achieving certain impacts through research use or trauma language. Therefore, the current work uses a relatively generous definition of research use, inclusive of claims that appear to draw from data or scientific knowledge or processes even if the source or validity of claims remains unclear. Like much research in this field, this work does not address the potential misuse of research evidence (Nutley et al., 2007).

Future Directions

Corresponding with the limited generalizability of the current work, additional in-depth analysis of legislation pertaining to other issue areas would allow comparisons to be made regarding how research may be leveraged differently depending on the issue area and the maturity of a relevant research base. Additionally, further refinement and validation of research keywords (e.g., Neal, Lawlor, Watling Neal, Mills, & McAlindon, 2018) would enable the use of big data on legislation for making comparisons of research use trends across issue areas such as health care, social services, and criminal justice. This work provided a first step for validating such keywords; however, the findings are limited to human trafficking legislation and could be expanded to examine use in all policy areas. Moreover, additional testing of keywords should consider whether the bill length or scope creates an upward bias of the estimated research use because expansive legislation containing multiple distinct programs (e.g., the Affordable Care Act) may be more likely to leverage research as an artifact of including many divergent or comprehensive components. Another validation check might examine whether individual legislators who make more public statements (e.g., during debate or messages to constituents) pertaining to what is “evidence based,” “best practice,” or effective are more likely to support bills using more research keywords. Finally, the use of machine and natural learning processing methods could be applied to full-text legislation datasets in order to improve the efficiency of content analysis (Li, 2016).

Conclusion

The movement toward evidence-based policy presents an opportunity for advancing trauma-responsive systems and trauma-informed services. There are numerous ways to leverage research in legislative language, including ways that implicitly support research-based practices by requiring evaluation or evidence review processes that define standards of “best practice.” Trauma language could also be leveraged explicitly in legislation when describing the problem and justifying action for specific policy proposals; however, the current work suggests that research language may be a particularly advantageous way of framing policy efforts.

Highlights.

• The evidence-based policy movement may present an opportunity for increasing the use of trauma-informed practices.

• The use of research evidence in human trafficking legislation, including references to trauma, has increased over time.

• Legislation leveraging research methods or findings has the potential to implicitly support trauma-informed care, especially because legislation that uses research language may be more likely to become enacted than legislation using trauma language.

• Policies that require research processes such as program evaluation and evidence reviews to guide service delivery standards have the potential to support trauma-informed practices in ways that are consistent with and adapt to the most current, evolving empirical base.