County-level Correlates of Human Trafficking Arrest Levels in Ohio: Social Disorganization,Social Capital,and Physical Contexts

Abstract

The present study investigated county-level correlates of human trafficking arrest levels in Ohio. Study variables were comprised of measures derived from social disorganization, social capital, and physical contexts of Ohio counties (N = 88). A negative binomial regression analysis was conducted to examine the relationship between county arrest counts and independent variables. Larger counts of human trafficking arrests were explained by higher levels of racial/ethnic heterogeneity, a social disorganization measure. Additionally, an increase in demand reduction strategy use was associated with a predicted increase in human trafficking arrest count. Further research on the influence of social variables and anti-human trafficking efforts on human trafficking arrest levels is needed to better understand how to effectively identify and combat human trafficking.

Keywords

human trafficking social disorganization social capital demand reduction anti-human trafficking efforts

Introduction

Human trafficking has been framed as a social problem regarding human rights violation, crime, and national security in the U.S. (Farrell & Fahy, 2009). Because of increased public and political awareness, human trafficking was established as a federal crime in the U.S. under the Trafficking Victims Protection Act (TVPA) of 2000. According to the TVPA (2000), severe forms of trafficking in persons occur when force, fraud, or coercion is used to subject individuals to commercial sexual exploitation or involuntary servitude. Means for prosecuting human traffickers, preventing human trafficking, and protecting human trafficking victims were also proposed within the TVPA as guidance for a national response to human trafficking. Part of these means include the efforts of federal and state agencies toward initiating human trafficking investigations and arresting suspects. In 2019, multiple federal agencies, including the Department of Defense and Department of Homeland Security, collectively initiated over 1,800 human-trafficking related cases or investigations (United States Department of State, 2020). Furthermore, the Federal Bureau of Investigation arrested 350 human trafficking suspects, and the Department of Homeland Security facilitated 2,197 arrests in 2019 (United States Department of State, 2020).

Human trafficking cases have garnered increased attention and bolstered anti-trafficking efforts from state agencies and researchers (Anderson et al., 2019; Office of the Attorney General, 2020; Ohio Human Trafficking Task Force, 2019a). States have begun to follow federal precedent regarding investigations into, and arrests for, human trafficking. Ohio was one of the last states to recognize human trafficking as a criminal offense by passing state legislation in 2010, but has made budding efforts toward identifying victims, prosecuting traffickers, and preventing trafficking over the past decade (Ohio Human Trafficking Task Force, 2019a, 2019b). In the State of Ohio, human trafficking is defined as the recruitment, luring, enticement, isolation, harboring, transportation, provision, obtainment, or maintenance of another person when the offender knows that the other person will be compelled to engage in sexual activity for hire or subjected to involuntary servitude (ORC 2905.32 [A][1]). The first human trafficking arrest in Ohio was made in 2011¹ under the newly defined state legislation (Office of Criminal Justice Services, 2020a).

The extensive interstate systems within Ohio have been proposed by researchers as a key correlate to the high human trafficking levels in the state (Davis, 2006). Recent research supports the relationship between interstate exit presence or proximity and human trafficking levels in other states (Diaz et al., 2022; Huff-Corzine et al., 2017; Mletzko et al., 2018). In addition to physical contexts (e.g., interstate exit presence), the presence of organized anti-trafficking efforts have also been recognized as key structural correlates to identify human trafficking. Research suggests that the use of human trafficking demand reduction strategies and participation in a human trafficking task force are positively associated with human trafficking identification levels (Diaz et al., 2022; Huff-Corzine et al., 2017). This relationship may be facilitated by the increased effort expended toward identifying victims and/or perpetrators by anti-trafficking groups (Diaz et al., 2022; Huff-Corzine et al., 2017). Anti-trafficking strategies have not yet been tested as correlates of human trafficking identification in Ohio, however, there have been efforts to measure the scope of trafficking victimization (Anderson et al., 2019) and improve data collection systems (Anderson et al., 2022). State government has prioritized the recognition of and solution to the human trafficking problem within Ohio (Exec. Order No. 2021-02D, 2021).

While examining human trafficking through physical and structural contexts is beneficial for understanding factors that contribute to the trafficking problem, exploration of social and community-level indicators of human trafficking may facilitate development of more efficient recognition of and response to trafficking in the U.S. (Huff-Corzine et al., 2017; Orme & Ross-Sheriff, 2015). Scholars have tested social processes that impact human trafficking on a community level; social disorganization theory has been applied to human trafficking arrest and offense counts at the neighborhood, county, and multi-national level, which has broadened the scope of human trafficking as a social problem (Diaz et al., 2022; Huff-Corzine et al., 2017; Mletzko et al., 2018). Indeed, extant work has found that two measures of social disorganization—concentrated disadvantage and residential instability—are significant indicators of human trafficking (Jiang & LaFree, 2017; Mletzko et al., 2018).

To further expand the body of research on social processes and factors as they influence the distribution of human trafficking arrests, this study will examine social disorganization variables as correlates of human trafficking arrest counts in Ohio counties (N = 88). Moreover, indices of social capital—community social groups and engagement that improve collective efficacy (Rupasingha et al., 2006)—were included as correlates of human trafficking arrest levels. The inclusion of social capital as an indicator of human trafficking is relatively unprecedented, so findings from this study may invite future research questions regarding the potential influence of social capital, which should influence levels of trust and cohesion, on human trafficking.

Human Trafficking and Social Disorganization

According to social disorganization theory, criminal patterns and social problems are spatially distributed throughout a geographical area based on structural characteristics of that area (Shaw & McKay, 1942). Social disorganization theory suggests that varying crime and delinquency rates among communities can be traced to different levels of uniformity regarding the morals and attitudes of members of these communities. Furthermore, Shaw and McKay (1942) proposed that three structural elements influence the social disorganization of a community: concentrated disadvantage, residential mobility, and racial/ethnic heterogeneity. High levels of these three determinants may be indicative of unsuccessful assimilation within a community (Bursik & Grasmick, 1993), which may discourage the formation of strong social ties among community members. Therefore, if a neighborhood possesses substantial amounts of concentrated disadvantage, residential mobility, and racial/ethnic heterogeneity, the development of neighborhood social networks is hindered, thus inhibiting the social organization needed to suppress crime. The inhibition of social organization may facilitate an environment that fosters criminal activity, including human trafficking.

There is empirical support for the effect of social disorganization components—and informal social control—on community crime levels. Early studies examining community characteristics (most often using city- or neighborhood-level crime and survey data) have found support for the general social disorganization framework as it applies to personal crimes (e.g., street robbery) and property crimes (e.g., burglary, motor vehicle theft, vandalism; Bellair, 1997; Sampson & Groves, 1989). Recent studies have also provided support for the ability of social disorganization theory to explain spatial variations in violent and property crimes committed by youth (Wong, 2012).

Studies featuring varying units of analysis have examined the potential relationship between social disorganization components and human trafficking. An analysis of structural correlates of human trafficking levels in Florida counties has been conducted (Diaz et al., 2022), as have analyses of spatial patterns of sex trafficking at the neighborhood level (Mletzko et al., 2018). Scholars have also investigated structural predictors of human trafficking on a multinational level (Jiang & LaFree, 2017; Karakus, 2008). Collectively, these studies yield support for the relationship between the spatial distribution of human trafficking and structural correlates that influence the ability of geographical areas to implement some form of social control. However, these studies did not use identical conceptualizations, definitions, and operationalizations of both human trafficking and social disorganization/social control variables and measures, so direct comparisons should be made with caution. For example, Jiang and LaFree (2017) included four types of trafficking within their definition of human trafficking (i.e., sexual exploitation, forced labor, removal of organs, illegal adoption), whereas the other studies included one or two types (e.g., sex trafficking, labor trafficking; Mletzko et al., 2018).

Jiang and LaFree (2017) concluded that lower levels of social control are associated with higher levels of human trafficking, and Karakus (2008) found that concentrated disadvantage—but not residential mobility and racial heterogeneity—is positively related to the cross-national distribution of human trafficking. Mletzko et al. (2018) identified that concentrated disadvantage was the sole significant predictor of human trafficking distribution. These findings suggested that communities characterized by higher levels of concentrated disadvantage may experience greater volumes of human trafficking due to the adverse effect of concentrated disadvantage on social control capability. Diaz et al. (2022) did not find significant relationships between the social disorganization measures and human trafficking arrest counts at the county level in Florida. However, further analysis of social disorganization and trafficking arrest counts at the county level in different contexts (i.e., replication) may facilitate better understanding of which social disorganization components—if any—influence county-level trafficking arrest distribution.

Collective Efficacy and Social Capital

Informal social control, influenced by concentrated disadvantage, residential mobility, and racial/ethnic heterogeneity, is thought by some scholars to influence criminal activity (Kornhauser, 1978; Sampson & Groves, 1989). The mechanism by which these elements of social disorganization impact crime distribution may be driven by collective efficacy, which is the social cohesion among community members that promotes reinforcement of acceptable—and expected—values and behaviors within the community (Sampson et al., 1997). Prior research supports the negative relationship between collective efficacy and crime levels (Bellair, 1997; Sampson & Groves, 1989).

Social capital is described by Coleman (1988) as an array of different entities that, collectively, bear some element of social structures and facilitate some form of action/activity for individuals within those social structures. Putnam (1995) characterized social capital as a facet of social organization (e.g., community social cohesion) that encourages in-group collaboration. Social capital can be obtained through civic engagement or other social participation in community-based organizations/activities (e.g., membership to community sport centers or teams). If the level of social participation and investment—and therefore, social capital—increases in a community, the ability of a community to implement informal social control inhibiting criminal activity should also increase (Putnam, 1995).

Prior studies have found ample support for the relationship between social capital and crime levels; Kennedy et al. (1998) found a significant negative relationship between social capital and firearm violence, and Rosenfeld et al. (2001) found a negative association between social capital levels and homicide rates. Finally, Deller and Deller (2010) found a significant relationship between robbery rates and overall social capital (while some specific measures of social capital were found to be related to different types of crime). Moreover, each of these studies suggest that social cohesion—and the economic impact of that cohesion—plays a crucial role in the distribution of various crime types throughout communities.

Human traffickers tend to exploit vulnerable populations for recruitment into trafficking (Cameron & Newman, 2008), and vulnerability may stem from many sources, including disadvantageous economic or social circumstances. Furthermore, the integrative impact of social and economic conditions as it relates to human trafficking levels throughout geographical areas should be explored, because identifying socioeconomic roots of conditions favorable to human trafficking may inform more effective anti-trafficking efforts (Williamson, 2019).

Human Trafficking Prevalence and Anti-Trafficking Efforts in Ohio

The State of Ohio is recognized as having a severe human trafficking problem by scholars (Davis, 2006; Perdue et al., 2010) and state departments (Office of the Attorney General, 2020). From 2015 to 2019, the fourth- or fifth-highest annual counts of requests on human trafficking cases involved a reference to Ohio (National Human Trafficking Hotline, 2015–2019). Anderson et al. (2019) illuminated the extent of human trafficking in Ohio by evaluating data collected from state agencies, service providers, and newspapers. Results of this Ohio prevalence study estimated that the count of known human trafficking victims from 2014 to 2016 was between 970 and 1,097 individuals. Additionally, the count of at-risk individuals during this time frame was estimated to fall within 4,083 to 4,338 cases (Anderson et al., 2019). Definitions of at-risk varied by data source but generally involved indications of shared vulnerabilities such as chronic running away and child sexual abuse (see Anderson et al., 2022 for detailed discussion of conceptual and operational definitions). These estimates were considered conservative by the authors, which they partially attributed to inconsistencies in data collection and sharing among anti-trafficking agencies. Variation in definitions and data capacity are two factors that hinder integrating agency data in Ohio (Anderson et al., 2022). Missing information, non-sharable data, and a lack of reaching out to many human trafficking victims (by social service or legal agencies) contribute to modest prevalence estimates of victims in Ohio (Anderson et al., 2022). However, the prevalence estimates that exist—while conservative—still suggest that human trafficking is prevalent.

Human trafficking was made a felony offense in the State of Ohio in 2010 through the passing of Senate Bill 235 (and has since been amended several times). Soon after passing human trafficking legislation, a state-level human trafficking response was developed in Ohio in 2012. This task force is grounded upon framework modeled by the U.S. Department of State, as its primary goals are to prevent trafficking, protect victims, and prosecute offenders (Ohio Human Trafficking Task Force, 2019a, 2019b). Fourteen state agencies (e.g., Ohio Department of Health, Ohio Department of Public Safety) are part of this task force, and each of these agencies collaborate to combat human trafficking in Ohio through various strategic approaches (e.g., find how to better identify victims; facilitating law enforcement collaborations to investigate human trafficking crimes; Ohio Human Trafficking Task Force, 2019a, 2019b).

The Ohio Attorney General’s Office also developed a statewide anti-human-trafficking effort of its own by implementing the Human Trafficking Initiative (HTI) in 2019, whose goals (like those of the state Task Force) are to fight human trafficking by advocating for community-based collaboration, victim aid, and prosecution of traffickers and johns (Office of the Attorney General, 2020). The HTI team works closely with anti-trafficking entities to facilitate a cooperative approach to curbing the human trafficking problem in Ohio. In 2020, the HTI reported the identification of 95 suspected human traffickers and 148 possible victims—as well as 76 arrests and 18 criminal convictions (Office of the Attorney General, 2020).

Further, the Ohio Organized Crime Investigations Commission (OOCIC) partners with HTI team members to establish support and collaboration among local law enforcement agencies in Ohio (Office of the Attorney General, 2020). Part of this collaboration has resulted in the formation of OOCIC human trafficking task forces throughout several regions in Ohio, each of which is composed of one or more county- or local-level law enforcement agencies. Scholars have found empirical support for the significant relationship between human trafficking arrest counts and the presence of a human trafficking task force (Huff-Corzine et al., 2017).

In addition to human trafficking task force presence being evaluated in this study, the presence of an anti-human-trafficking coalition as it relates to human trafficking arrest counts at the county level is examined. A joint goal of the Ohio Attorney General’s Office and the Ohio Human Trafficking Task Force (OHTTF) is to increase collaboration among law enforcement agencies, service providers, and anti-trafficking groups/organizations. Furthermore, more Ohio counties are served by a regional anti-trafficking coalition than they are covered by an OOCIC task force. Examining the potential effects of these disparities on human trafficking arrest levels may provide further insight into the role and efficiency of specific organized anti-trafficking efforts.

The Current Study

Overall, the current study adds to the growing body of work that examines physical and structural contexts as well as social and community indicators as correlates of human trafficking arrest counts to better assist counties, especially those in Ohio, in their response to human trafficking. There are three overarching goals. First, this study evaluates theoretically derived indicators of county-level social disorganization and social capital as it relates to human trafficking arrest counts. Second, analyses account for structural characteristics, including anti-trafficking efforts. Finally, we attempted, as closely as possible, to replicate the methodological approach and findings of Diaz et al. (2022) analysis of the structural determinants of human trafficking arrests across Floridan counties through an analysis of Ohio counties.

Methods

Data Description

The crime data featured in this study was accessed through the Ohio Incident Based Reporting System (OIBRS). The OIBRS is modeled after the FBI’s National Incident-Based Reporting System (NIBRS), as it is a voluntary crime reporting system to which crime offense information can be uploaded by law enforcement agencies in a standardized structure (Office of Criminal Justice Services, 2020a, 2020b). This crime reporting system has facilitated more accurate reporting of criminal offenses, because the NIBRS collects information for up to 10 criminal offenses during an incident and includes data elements regarding various factors within a crime incident, such as offender information, arrestee information, and victim information (Office of Criminal Justice Services, 2020a, 2020b). Human trafficking arrests can be recorded as “commercial sex acts” (i.e., sex trafficking) and/or “involuntary servitude” (i.e., labor trafficking) within OIBRS, however these specifications are infrequently reported into OIBRS by law enforcement agencies. OIBRS also includes extended codes for data elements (e.g., location type, weapon type; Office of Criminal Justice Services, 2020a, 2020b). The crime incident detail collected per OIBRS protocol facilitates advantageous data collection—including the collection of county-level arrest data utilized within this study. In addition to the OIBRS, several sources providing information at the county-level were employed for data collection. Data were collected through the American Community Survey (ACS; United States Census Bureau, 2021a) and state agencies (Office of the Attorney General, 2020; Ohio Human Trafficking Task Force, 2019a, 2019b). Data were also derived from databases created by researchers on social capital (Rupasingha et al., 2006) and sex trafficking demand reduction (Shively et al., 2012). The analytic sample was 88 counties.

Measures

Dependent variable

Human trafficking arrests

Human trafficking was deemed a felony offense in Ohio in 2010. The first arrest related to human trafficking was recorded in the OIBRS in 2011, so human trafficking OIBRS arrest data collected from 2011 to 2018 is included for analysis. There were 69 total human trafficking arrests reported to the OIBRS during this time frame. However, 11 of these arrests were not attributed to a specific Ohio county. Because Ohio counties are the unit of analysis, and the count of trafficking arrests for each Ohio county serves as the dependent variable, these 11 arrests were excluded. The resulting count of human trafficking arrests included for analysis was 58.

Independent variables

Prostitution-related arrests

Since human trafficking crimes are commonly treated by the criminal justice system as prostitution-related crimes (Farrell et al., 2014), the counts of prostitution-related arrests for each county from 2011 to 2018 were also collected from the OIBRS. Because OIBRS data feature up to 10 separate offenses within each crime incident, some incidents were associated with multiple embedded prostitution-related offenses. Each separate offense was counted toward the number of prostitution arrests per county. Prostitution-related arrests included offenses of compelling prostitution (i.e., attempt to encourage, convince, or facilitate a minor’s engagement in sexual activity), promoting prostitution (e.g., offender supervised or managed a prostitute or brothel or aided the transportation of prostitutes across state lines), procuring prostitution (e.g., facilitated engagement between a prostitute and direct patron or utilized premises for prostitution), and soliciting prostitution (e.g., hired an individual for sexual activity with some form of payment; Office of Criminal Justice Services, 2020a, 2020b).

Social disorganization

One of the key theoretical frameworks tested in this study is social disorganization theory. Data representing these attributes were obtained from the ACS 5-year estimates (2014–2019) produced by the United States Census Bureau (2021b). Indices of concentrated disadvantage, residential instability, and racial/ethnic heterogeneity were included (Bursik & Grasmick, 1993; Diaz et al., 2022; Huff-Corzine et al., 2017; Mletzko et al., 2018).

The operationalization and coding of all three social disorganization indices were modeled after previous scholars (Bursik & Grasmick, 1993; Diaz et al., 2022; Huff-Corzine et al., 2017; Mletzko et al., 2018). The theoretically derived concentrated disadvantage index was comprised of five values (α = .44), each of which were standardized and then averaged to create a final concentrated disadvantage score. The following items comprised the concentrated disadvantage index: the percent of families below the poverty level; the percent of households receiving foods stamps/SNAP; the percent of female-only headed households with children under 18 years of age; the percentage of people in the labor force unemployed; and the percentage of the population aged 25 years and over with a bachelor’s degree. Each of these five items—reflecting levels of poverty, unemployment, and educational attainment—comprised the concentrated disadvantage variables utilized within previous studies identifying significant effects of concentrated disadvantage on human trafficking identification distribution (Diaz et al., 2022; Mletzko et al., 2018).

Residential instability was also calculated by standardizing each sub-value and averaging the standardized scores to create a residential instability index. Two values (α = .74) comprised the residential instability scale: the percent of renter-occupied units and the percent of the population living in a different household from the previous year.

The final social disorganization variable—racial/ethnic heterogeneity—constituted seven racial/ethnic categories: non-Hispanic whites; non-Hispanic blacks; non-Hispanic Asian; non-Hispanic Native Hawaiian and Other Pacific Islanders; non-Hispanic other races; and Hispanics. The sum of the squared proportions of each category was used to create the racial/ethnic heterogeneity index for each county, ranging from 0 (full heterogeneity) to 1 (full homogeneity). Values were subtracted by 1 so that higher values indicated higher levels of heterogeneity. Scores were standardized to facilitate comparison of effects of social disorganization variables on human trafficking arrests counts. These indices were calculated based on a measure created by Blau (1977) and later used by other researchers investigating human trafficking levels through a social disorganization framework (Diaz et al., 2022; Mletzko et al., 2018).

Social capital

Indices of social capital for all 88 counties were collected from a nationwide measure of social capital created by Rupasingha et al. (2006). The social capital indices were calculated based on national data retrieved from the County Business Patterns dataset—produced by the U.S. Census Bureau. Four standardized variables were used to create the social capital index via principal component analysis. The first primary variable consisted of 10 county-level values including the number of religious organizations; civic and social organizations; business associations; political organizations; professional organizations; labor organizations; bowling centers; fitness and recreational sports centers; golf courses and country clubs; sports teams and clubs. Each of these items can be considered groups in civil society that present county-level opportunity for associational membership and enhance community interaction (Rupasingha et al., 2006). The additional primary variables were comprised of voter turnout, census response rate, and number of non-profit organizations without including those with an international approach (Rupasingha et al., 2006). These four primary variables comprised the social capital index, which was also standardized.

Demand reduction strategies

The number of demand reduction strategies present within each county as of 2020 was obtained from Demandforum.org, a web source currently impelled by the National Center on Sexual Exploitation (NCOSE). This source was most recently updated in 2020, and it allows for the navigation of all county-level programs and methods employed to curb demand for sex trafficking or prostitution (Shively et al., 2012). The 12 following demand reduction efforts are available for evaluation on the Demandforum.org website: auto seizure; cameras; community service; John school; letters; license suspension; neighborhood action; public education; reverse stings; shaming; “Stay Out of Areas with Prostitution” (SOAP) orders; and web stings. The count of demand reduction strategies implemented within each county as of 2020 is included as the demand reduction variable.

Law enforcement agencies

A law enforcement directory provided by the Ohio Office of the Attorney General was used to compile the count of law enforcement agencies within each county. These agencies included local/organizational police departments, county sheriff’s offices, and specialized police agencies (e.g., those located within metro parks). The count of law enforcement agencies in a county or geographical area has been recognized and utilized as a proxy measure for population (Diaz et al., 2022; Jurek & King, 2020), so law enforcement agency count is also considered a proxy measure for county population.

Participation in human trafficking task force

Participation in a human trafficking task force was determined based on whether counties provided law enforcement resources toward one of the three currently operating human trafficking task forces under the OOCIC. This included any county containing at least one law enforcement agency credited with participation in a state human trafficking task force. To facilitate the inclusion of this measure within the selected statistical model used in this study, participation in a task force was coded as a binary variable (1 = yes; 0 = no).²

Analytic Strategy

Descriptive statistics were calculated and reported for each variable (Table 1). Bivariate relationships and multicollinearity levels were then assessed through the computation of Pearson correlations among all variables (Table 2). All predictor variables in the model garnered VIFs below four, which adhered to the rule of thumb delineated by Yu et al. (2015).

Table 1.

Descriptive Statistics for all Variables (N = 88).

Variable	Mean/%	SD	Min	Max
Human trafficking arrest count	0.66	2.82	0	20
Prostitution-related arrest count	127.70	697.02	0	6,306
Concentrated disadvantage index (std)	0	0.656	−1.59	1.53
Residential instability index (std)	0	0.92	−1.95	3.85
Racial/ethnic heterogeneity index (std)	0	1.00	−1.22	3.30
Social capital index (std)	−.11	0.46	−1.36	0.92
Demand reduction strategy count	0.51	1.16	0	5
Law enforcement agency count	11.15	11.16	1	76
Participation in task force	6.8%	0.5	0	1

Table 2.

Bivariate Pearson Correlations of All Study Variables (N = 88).

	A	B	C	D	E	F	G	H	I
A. Human trafficking arrest count	1.00
B. Concentrated disadvantage (std)	.250*	1.00
C. Residential instability (std)	.232*	.572**	1.00
D. Racial/ethnic heterogeneity (std)	.444**	.470**	.591**	1.00
E. Social capital (std)	−.063	−.284**	−.171	−.030	1.00
F. Demand reduction strategy count	.454**	.384**	.333**	.567**	−.114	1.00
G. Prostitution-related arrest count	.180	.250*	.386**	.489**	−.027	.288**	1.00
H. Law enforcement agency count	.482**	.440**	.454**	.775**	−.074	.561**	.553**	1.00
I. Participation in task force	.130	.275**	.125	.412**	.002	.311**	.472**	.541**	1.00

Correlation is significant at the .05 level (two-tailed).

Correlation is significant at the .01 level (two-tailed).

The dependent variable is represented by count data—specifically, the count of human trafficking arrests—and these count data were found to be over-dispersed (M = 0.66 to SD = 2.82). Due to the over-dispersion of this count data, a negative binomial regression model was deemed the best fit. The influence of social disorganization, social capital, anti-trafficking efforts, prostitution-related arrests, and law enforcement agency count on human trafficking arrest counts were evaluated with the model. Ohio counties represented the unit of analyses.

Results

Descriptive Statistics

Descriptive statistics of all study variables are displayed below in Table 1. From 2011 to 2019, there was a total of 58 human trafficking arrests recorded in the OIBRS that can be attributed to a single Ohio county. Out of all 88 Ohio counties, only 16% were associated with at least one human trafficking arrest. The highest count of human trafficking arrests derived from Summit County (20 arrests), followed by Montgomery County (15 arrests) and Cuyahoga County (9 arrests).

Because all social disorganization measures were calculated as standardized indices, mean scores for concentrated disadvantage (Range = −1.59 to 1.53), residential instability (−1.95 to 3.85), and racial/ethnic heterogeneity (Range = −1.22 to 3.30) were all 0. Finally, the social capital index (M = −0.11 to SD = 0.46) yielded the smallest range out of all indices included in the study (Range −1.36 to 0.92). The average count of human trafficking arrests per county was relatively low (M = 0.66) compared to the level of prostitution-related arrests per county (M = 127.70). Only 16 Ohio counties used at least one demand reduction strategy, and the counties with the highest recorded use of demand reduction strategies implemented five (Lucas County) and four (Montgomery County) strategies. Approximately 6.8% of Ohio counties participated in a human trafficking task force. The number of police agencies per county yielded a wide range, as the minimum police agency count in a county was 1, and the maximum count was 76.

Bivariate Correlations

Bivariate statistics are reported in Table 2. Statistical tests revealed several significant relationships between the dependent variable and independent variables featured in this study. As anticipated, there were significant relationships between human trafficking arrest count and each of the social disorganization variables. The correlation between human trafficking arrest number and racial/ethnic heterogeneity (r(86) = .44 to p < .01) was moderate, while small bivariate correlations were found between the dependent variable and both concentrated disadvantage (r(86) = .25 to p < .05) and residential instability (r(86) = .23 to p < .05). Unexpectedly, there was no significant correlation between human trafficking arrest count and the social capital measure.

Human trafficking arrest count and both prostitution-related arrest count and participation in a human trafficking task force were not significantly correlated. Law enforcement agency count, however, was moderately correlated with the human trafficking arrest count (r(86) = .48 to p < .01), as was the demand reduction strategy count (r(86) = .45 to p < .01). Most of the correlations found among the predictors were weak or moderate, though there was a strong correlation identified between racial heterogeneity and law enforcement agency count, (r(86) = .78 to p < .01). These findings suggest that higher levels of human trafficking arrests are associated with higher levels of social disorganization, demand reduction strategy use, and police agency count.

Negative Binomial Regression

Table 3 displays findings from the negative binomial regression. The omnibus test was significant (likelihood ratio x² (8) = 90.24 to p < .001), indicating that at least one predictor variable included in the model influences the levels of human trafficking arrests throughout Ohio counties. The model including this study’s predictor variables represents a significant improvement in fit compared to the null model without any independent variables. As anticipated, larger counts of human trafficking arrests were explained by higher levels of racial/ethnic heterogeneity and increased use of demand reduction strategies. All remaining predictor variables included in the model were not significant.

Table 3.

Negative Binomial Regression Model Predicting Human Trafficking Arrests (N = 88).

Variable	Coef.	IRR	Std. Err.	Sig.
Concentrated disadvantage index (std)	−1.42	0.24	0.82	.082
Residential instability index (std)	0.13	1.14	0.47	.789
Racial/ethnic heterogeneity index (std)	1.31	3.72	0.49	.007**
Social capital index (std)	−0.65	0.52	0.81	.425
Demand reduction strategy count	0.49	1.64	0.23	.029*
Prostitution-related arrest count	−0.00	1.00	0.00	.122
Law enforcement agency count	0.04	1.04	0.04	.221
Participation in task force	−0.19	0.83	0.92	.840

Note. Std. Err. is presented for the Coef. IRR = incidence rate ratio.

Relationship is significant at the .05 level (two-tailed).

Relationship is significant at the .01 level (two-tailed).

The incidence rate ratios (IRRs) reported in Table 3 reflect the changes in incident rates of human trafficking arrests for each one unit increase of every predictor variable included in the model. These IRRs also represent a relative effect size measure of the independent variables on the count of human trafficking arrests recorded in the OIBRS by Ohio counties from 2011 to 2018. Every standardized unit increase in the racial/ethnic heterogeneity index was associated with a predicted 272% increase in the count of human trafficking arrests. The number of demand reduction strategies employed by each county also garnered a significant IRR (p < .05). The negative binomial regression model estimated a 64% increase in human trafficking arrests for every additional (single) demand reduction strategy implemented within Ohio counties. These findings suggest that racial/ethnic heterogeneity and demand reduction strategies are integral facets to the explanation of different human trafficking arrest levels at the county level. Aside from racial/ethnic heterogeneity and demand reduction strategy count, other predictor variables that garnered moderate relationships with human trafficking arrest count were concentrated disadvantage index (IRR = .241 to p = .082) and social capital (IRR = 0.522 to p = .425). However, these effect sizes were not significant.

Discussion

The objective of this study was to better understand human trafficking arrest patterns in Ohio and to recognize county level variables that may influence the distribution of those arrests throughout space. High levels of concentrated disadvantage, residential instability, and racial/ethnic heterogeneity may inhibit communities from enforcing productive social control, theorized to increase criminal activity (Bellair, 1997; Kornhauser, 1978; Sampson & Groves, 1989; Sampson et al., 1997). This study featured predictor variables that were modeled based on the work of Diaz et al. (2022; i.e., concentrated disadvantage, residential instability, racial/ethnic heterogeneity, demand reduction strategy count, law enforcement agency count, participation in a task force, prostitution-related arrest count). Social capital variables examined relative to human trafficking arrest count, however, was a novel addition to the current study. Whereas support for social disorganization indicators to predict variations in human trafficking levels has been found since the establishment of the TVPA (Diaz et al., 2022; Huff-Corzine et al., 2017; Karakus, 2008; Mletzko et al., 2018), the current study found only one of those indicators—racial/ethnic heterogeneity—was associated with human trafficking arrests in the state of Ohio.

This study further explored the capacity of social disorganization correlates to explain variations in human trafficking by examining human trafficking arrests and structural contexts at the Ohio county level. Social capital—a facet of social organization that encourages collaboration and collective efficacy—was also introduced as a potential predictor of human trafficking arrest levels, grounded in previous research supporting the association between social capital scores and distribution of other crime types (Deller & Deller, 2010; Kennedy et al., 1988; Rosenfeld et al., 2001). In addition to examining the ability of social disorganization and social capital variables to explain variations in the count of human trafficking arrests across Ohio counties, prostitution-related arrest counts, count of law enforcement agencies, and anti-trafficking efforts (i.e., demand reduction strategy use, participation in a task force) were assessed as predictors of human trafficking arrest levels.

Findings from the negative binomial regression revealed that at least one social disorganization variable (i.e., racial/ethnic heterogeneity) and one of the anti-trafficking effort variables (i.e., demand reduction) were significant predictors of human trafficking arrest counts. This study yielded support for the notion that human trafficking arrests are not randomly distributed throughout Ohio counties. Furthermore, these findings provide partial support for the application of social disorganization theory to the crime of human trafficking, which adds to the growing body of research regarding human trafficking and social disorganization. Interestingly, the discovery of racial/ethnic heterogeneity as a significant predictor of human trafficking arrest counts (in the expected predictive direction) is a relatively new finding compared to previous work conducted on social disorganization and human trafficking (Diaz et al., 2022; Huff-Corzine et al., 2017; Karakus, 2008; Mletzko et al., 2018). The results of this study also suggest that increased use of demand reduction strategies is associated with higher human trafficking arrest counts, which aligned with the expectation that counties garnering increased human trafficking arrests may exert more effort to identify and combat human trafficking in that county.

While racial/ethnic heterogeneity and demand reduction strategy count were found to be significant predictors of human trafficking arrest counts, the remaining predictors included in the statistical model did not yield significant relationships with human trafficking arrests. Two of the three included social disorganization variables (concentrated disadvantage,³ residential instability) were nonsignificant. This departs from prior work that has found both concentrated disadvantage and/or residential instability as significant predictors of human trafficking levels (Diaz et al., 2022; Huff-Corzine et al., 2017; Karakus, 2008; Mletzko et al., 2018). The effect of these variables on human trafficking distribution should continue to be explored.

The application of social capital to human trafficking levels is relatively new. Social capital yielded a nonsignificant—though moderate—incidence rate ratio in this study. Additionally, a supplemental negative binomial regression analysis was conducted to investigate the four factors comprising the social capital measure as they may relate to human trafficking arrest counts throughout Ohio counties. One factor—the number of non-profit organizations without including those with an international approach—yielded a significant IRR (IRR = 1.001 to p < .001). This suggests that some aspects of social capital—specifically, the aspects representing economic and social prosperity—may be key elements to understanding human trafficking identification distribution.

In addition to providing support for the effect of racial/ethnic heterogeneity and demand reduction strategies on human trafficking arrest distribution, the findings from these analyses align with prior research regarding the significance of economic deprivation and low social ties among communities as indicators of human trafficking. Prior research has found the concentrated disadvantage and residential instability of citizens or households as significant correlates of human trafficking, supporting social disorganization theory as it applies to human trafficking (Diaz et al., 2022; Huff-Corzine et al., 2017; Karakus, 2008; Mletzko et al., 2018). The significant (and non-significant—but moderate) IRRs yielded by the social capital factors indicated that the presence of economically and socially prosperous organizations within a county (or community) may also contribute to a geographical area’s vulnerability to human trafficking. Moreover, further examination of social capital as it relates to human trafficking may better researchers’ understanding of its role as a correlate of human trafficking identification.

The nonsignificant findings regarding the physical correlate variable (i.e. count of law enforcement agencies) included in the model suggest that more fluid factors (such as social or economic factors and anti-trafficking efforts) may be more influential in explaining human trafficking distribution than are concrete factors. Additionally, the relatively small and nonsignificant IRR yielded by participation in a human trafficking task force could be reflective of a low number of OOICIC task forces that were considered for measurement (only three currently exist within the State of Ohio). Presence or participation in task forces has been found as a significant predictor of higher human trafficking identification levels by some researchers in the past (Huff-Corzine et al., 2017), though not by all (Diaz et al., 2022), so continued exploration of the relationship between task force participation and human trafficking arrests could aid understanding of how human trafficking identification is distributed.

Policy and Practice Considerations

The findings of this study may promote several key policy recommendations regarding the recognition and combatting of human trafficking at the county level. First, because counties utilizing more counts of demand reduction strategies were generally associated with higher human trafficking arrest levels, the implementation of additional demand reduction strategies throughout the state may facilitate higher identification of human trafficking offenses. Adding resources toward existing demand reduction tactics (to increase their efficacy) and invoking the use of demand reduction techniques in counties that do not already utilize them may further improve human trafficking identification in Ohio.

In addition to implementing more demand reduction strategies, statewide efforts to improve the social and economic conditions of high-trafficking counties may obstruct the flow of human trafficking in vulnerable communities. Policies aimed at facilitating productive racial integration or diminishing racial biases among communities may improve the social organization of high-trafficking areas, as might policies designed to bridle the economic disadvantage of community members. This may be accomplished through community policing strategies or might be a more global approach that recognizes and addresses the underlying socio-structural forces (e.g., systemic racism) that permutate the legal system and exacerbate criminal justice system contact. Moreover, the results of the follow-up analysis with all four social capital factors included in the model suggested that the volume of organizational opportunities and involvement among counties may be linked to counts of human trafficking arrests. Obtaining government funding for the creation of more non-profit organizations and/or business, sports, social, or political organizations may improve the economic and social prosperity of counties (or communities) vulnerable to high levels of human trafficking and again echoes a community policing approach. Moreover, the development of social and economic prosperity may—hopefully—aid in the curbing human trafficking.

Limitations

While this study meaningfully contributes to the growing body of literature applying social disorganization theory to human trafficking levels, limitations of the study should be addressed. Due to the nature of the data collection of the OIBRS (voluntary reporting), there is a chance that the true count of human trafficking arrests in Ohio from 2011 to 2018 is higher than what was detailed in the OIBRS dataset. As more counties and law enforcement agencies adopt the OIBRS, values of human trafficking offenses may become more reflective of the “true” count of offenses known to law enforcement. Relatedly, the misclassification of human trafficking victims by law enforcement officials as offenders (of trafficking or other crimes) may contribute to the undercounting of human trafficking offenses in law enforcement reporting systems (Farrell et al., 2019). Moreover, victims may be overlooked entirely by law enforcement, resulting in a lack of arrests of human trafficking offenders and contributing to the hidden figure of crime (Farrell et al., 2010).

In addition to the potential underestimation of the count of human trafficking arrests (the dependent variable) within this study, one of the independent variables—participation in a human trafficking task force—may have also underrepresented the presence of task forces throughout Ohio counties. Only OOCIC human task forces were included as part of this measure due to issues in obtaining reliable sources documenting any other task forces. It is possible that other human trafficking task forces existed throughout Ohio, but they were not included within the measure if they were not affiliated with the OOCIC. Another independent variable—count of sex trafficking demand reduction strategies—may also be considered a limitation of this study. This measure reflects the number of strategies implemented within each Ohio county as of 2020, so it is not possible to discern how many of these strategies were implemented prior to, during, or after the time frame during which the human trafficking arrest data were collected.

Other limitations of this study are its generalizability to areas outside of Ohio and its sample size. The sample utilized in this study consisted of 88 counties, so results of this study should not be generalized to other states or other units of analysis within Ohio outside of the county level. By nature of counties, the sample size is small; this likely reduced the statistical power of the negative binomial regression analysis. Additionally, the data of the dependent variable (human trafficking arrest count) was zero-inflated, as many counties did not record any human trafficking arrest from 2011 to 2018. A negative binomial regression was deemed the best fitting model for the data—despite the limitations of using this model with a smaller sample size and seemingly zero-inflated dependent variable data. The results of the negative binomial regression analysis, however, indicate that the analyses conducted were sufficient to contribute to research regarding physical and structural correlates of human trafficking arrests.

Future Research Directions

Future research should investigate human trafficking arrest levels at the county level. Utilizing a county-level approach may facilitate comparisons across findings conducted in Ohio and across other states (e.g., Florida; Diaz et al., 2022). Further, researchers should investigate if state-level incident-based reporting data (e.g., OIBRS) includes greater human trafficking arrest coverage than what ends up reported into NIBRS. Given variation in law enforcement agency adoption of incident-based reporting and implementation of state-specific human trafficking legislation, there may be differences in state-level reporting compared to national reporting. Replicated use of social disorganization, social capital, and demand reduction strategies as indicators may contribute meaningfully to researchers’ understandings of how anti-trafficking efforts and macro-level economic and social factors are related to human trafficking identification distribution. The application of social disorganization and social capital theories is relatively new in human trafficking research (though not in research on other types of community crime)—but these applications are promising—so the relationships between these variables should continue to be explored. Furthermore, solidifying and agreeing upon a single definition of social capital may facilitate direct comparison across any future studies that examine the link between social capital and human trafficking levels. Utilizing uniform operationalizations of this concept may strengthen its position within the body of literature concerning human trafficking and economic and social conditions.

Conclusion

This study examined the influence of various physical and structural contexts on human trafficking arrest counts throughout Ohio counties. Partial support for the application of social disorganization theory to human trafficking levels was obtained, as was support for the impact of demand reduction strategy use on the count of human trafficking arrests. Though the remaining independent variables entered in the negative binomial regression model yielded nonsignificant effects on human trafficking arrest count, the findings of this study invite further discussion and exploration of the spatial relationship between human trafficking and structural variables. These findings also illuminate the need for continued research on malleable contributors to human trafficking distribution because these factors may—ideally—be improved/changed through anti-trafficking statutes or policies. Moreover, further evaluating the impact of social variables and anti-trafficking efforts on human trafficking levels may facilitate meaningful progression in the way we understand and combat human trafficking on a local, county, state, and national scale.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Valerie R. Anderson

Notes

Author Biographies

Hillary L. Benanzer, MS, received a bachelor’s degree in psychology from Ohio University in 2020, where she contributed research efforts to the fields of psychotherapy and human judgment and decision-making. She completed her master’s degree in criminal justice at the University of Cincinnati in 2021, with a concentration in law enforcement and crime prevention. Her recent research includes a scoping review of sociological approaches and human trafficking interventions.

Valerie R. Anderson, PhD, is an Assistant Professor in the School of Criminal Justice at the University of Cincinnati. The focus of her research program is to understand youth offending and victimization, and—more specifically—the circumstances and contexts in which these areas intersect. Her recent work is featured in American Journal of Public Health, American Journal of Community Psychology, Criminal Justice and Behavior, and Feminist Criminology.

Brittany E. Hayes, PhD, is an Assistant Professor in the School of Criminal Justice at the University of Cincinnati. Her research centers on victimization as well as how the broader social context influences individuals’ perceptions and behaviors. Her work has been published in the American Sociological Review, Justice Quarterly, and the Journal of Research in Crime and Delinquency.

References

Anderson

V. R.

Kulig

T. C.

Sullivan

C. J.

(2019). Estimating the prevalence of human trafficking in Ohio, 2014-2016. American Journal of Public Health, 109(10), 1396–1399.

Anderson

V. R.

Kulig

T. C.

Sullivan

C. J.

(2022). Integrating human trafficking data: A case study of conceptual and operational variation in Ohio. Journal of Human Trafficking, 8(3), 250–264.

Bellair

P. E.

(1997). Social interaction and community crime: Examining the importance of neighbor networks. Criminology, 35(4), 677–704.

Blau

P. M.

(1977). A macrosociological theory of social structure. American Journal of Sociology, 83(1), 26–65.

Bursik

Grasmick

(1993). Economic deprivation and neighborhood crime rates, 1960-1980. Law & Society Review, 27(2), 263–283.

Cameron

Newman

(2008). Trafficking in humans: Social, cultural and political dimensions. United Nations University Press.

Coleman

J. S.

(1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95–S120.

Davis

K. Y. S.

(2006). Human Trafficking and modern day slavery in Ohio. Polaris Project Ohio.

Deller

S. C.

Deller

M. A.

(2010). Rural crime and social capital. Growth and Change, 41(2), 221–275.

10.

Diaz

Huff-Corzine

Corzine

(2022). Demanding reduction: A county-level analysis examining structural determinants of human trafficking arrests in Florida. Crime and Delinquency, 68, 28–51. https://doi.org/10.1177/0011128720962710

11.

Exec. Order No. 2021-02D. (2021, January 29).

12.

Farrell

Dank

Kafafian

Lockwood

Pfeffer

Hughes

Vincent

(2019). Capturing human trafficking victimization through crime reporting. U.S. Department of Justice.

13.

Farrell

Fahy

(2009). The problem of human trafficking in the US: public frames and policy responses. Criminal Justice Journal, 37(6), 617–626.

14.

Farrell

McDevitt

Fahy

(2010). Where are all the victims? Understanding the determinants of official identification of human trafficking incidents. Criminology & Public Policy, 9(2), 201–233.

15.

Farrell

Owens

McDevitt

(2014). New laws but few cases: Understanding the challenges to the investigation and prosecution of human trafficking cases. Crime Law and Social Change, 61, 139–168.

16.

Huff-Corzine

Sacra

S. A.

Corzine

Rados

(2017). Florida’s task force approach to combat human trafficking: An analysis of county-level data. Police Practice and Research, 18(3), 245–258.

17.

Jiang

LaFree

(2017). Social control, trade openness and human trafficking. Journal of Quantitative Criminology, 33(4), 887–913. https://doi.org/10.1007/s10940-016-9316-7

18.

Jurek

A. L.

King

W. R.

(2020). Structural responses to gendered social problems: Police agency adaptations to human trafficking. Police Quarterly, 23(1), 25–54.

19.

Kennedy

B. P.

Kawachi

Prothrow-Stith

Lochner

Gupta

(1998). Social capital, income inequality, and firearm violent crime. Social Science & Medicine, 47(1), 7–17.

20.

Kornhauser

R. R.

(1978). Social sources of delinquency: An appraisal of analytic models. University of Chicago Press.

21.

Mletzko

Summers

Arnio

A. N.

(2018). Spatial patterns of urban sex trafficking. Criminal Justice Journal, 58, 87–96.

22.

National Human Trafficking Hotline. (2015–2019). Hotline statistics. https://humantraffickinghotline.org/states

23.

Office of Criminal Justice Services. (2020a). Ohio incident-Based Reporting System. Ohio Department of Public Safety. https://ocjs.ohio.gov/oibrs/index.aspx

24.

Office of Criminal Justice Services. (2020b). OIBRS data collection and submission specifications. Ohio Department of Public Safety. https://ocjs.ohio.gov/oibrs/links/Data_Specs.pdf

25.

Office of the Attorney General. (2020). Annual report: 2020 human trafficking. Author.

26.

Ohio Human Trafficking Task Force. (2019a). Governor’s Ohio human trafficking task force report: January 2019. Ohio Department of Public Safety, Office of Criminal Justice Services.

27.

Ohio Human Trafficking Task Force. (2019b). Map of Ohio’s anti-human trafficking coalitions. Ohio Department of Public Safety, Office of Criminal Justice Services. https://www.humantrafficking.ohio.gov/coalitions.html

28.

Orme

Ross-Sheriff

(2015). Sex trafficking: Policies, programs, and services. Social Work, 60(4), 287–294.

29.

Perdue

Williamson

Billings

Schart

Boston-Gromer

(2010). In the matter of human trafficking in Ohio: The pursuit of justice continues. Women’s Policy Journal of Harvard, 8, 3–12.

30.

Putnam

R. D.

(1995). Bowling alone: America’s declining social capital. Journal of Democracy, 6(1), 65–78. https://doi.org/10.1353/jod.1995.0002

31.

Rosenfeld

Baumer

E. P.

Messner

S. F.

(2001). Social capital and homicide. Social Forces, 80(1), 283–310.

32.

Rupasingha

Goetz

S. J.

Freshwater

(2006). The production of social capital in US counties. The Journal of Socioeconomics, 35(1), 83–101.

33.

Sampson

R. J.

Groves

W. B.

(1989). Community structure and crime: Testing social-disorganization theory. American Journal of Sociology, 94(4), 774–802. https://doi.org/10.1086/229068

34.

Sampson

R. J.

Raudenbush

S. W.

Earls

(1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277(5328), 918–924.

35.

Schwarz

Alvord

Daley

Ramaswamy

Rauscher

Britton

(2019). The trafficking continuum: Service providers’ perspectives on vulnerability, exploitation, and trafficking. Affilia, 34(1), 116–132.

36.

Shaw

McKay

(1942). Juvenile delinquency and urban areas. Science, 277(5328), 918–924. https://doi.org/10.1126/science.277.5328.918

37.

Shively

Kliorys

Wheeler

Hunt

(2012). A national overview of prostitution and sex trafficking demand reduction efforts, final report (Report No. 238796). U.S. Department of Justice. https://www.ojp.gov/pdffiles1/nij/grants/238796.pdf

38.

TVPA. (2000). Victims of Trafficking and Violence Protection Act of 2000, Pub. L. 106-386, 114 Stat. 1464, codified as amended at 22 U.S.C. §§ 7101-7110.

39.

United States Census Bureau. (2021a). American community survey data. https://www.census.gov/programs-surveys/acs/data.html

40.

United States Census Bureau. (2021b). State population totals: 2010-2020. https://www.census.gov/newsroom/press-releases/2020/estimates-population-housing-units.html

41.

United States Department of State. (2020). 2020 report on U.S. government efforts to combat trafficking in persons. United States Department of State.

42.

Williamson

S. H.

(2019). Toward an integrated theory of human trafficking: Unraveling the role of global economic policies in shaping macro-level conditions cross-nationally (Publication No. 27801578) [Doctoral Dissertation, North Carolina State University]. ProQuest Dissertations Publishing.

43.

Wong

S. K.

(2012). Youth crime and family disruption in Canadian municipalities: An adaptation of Shaw and McKay’s social disorganization theory. International Journal of Law Crime and Justice, 40, 100–114.

44.

Jiang

Land

K. C.

(2015). Multicollinearity in hierarchical linear models. Social Science Research, 53, 118–136.