Are Human Trafficking Severity Scores Comparable across Contexts? Evidence from Malawi and Zambia

Background and Purpose:
Social work researchers and practitioners increasingly use survey-based trafficking indicators to identify high-risk populations, estimate prevalence, and design prevention and intervention efforts. Yet these uses depend on a critical assumption: that trafficking indicators carry similar meaning across populations and settings. When indicators function differently across countries or sampling contexts, cross-group comparisons may misidentify who is most at risk, distort prevalence estimates, and weaken decisions about service targeting, resource allocation, and policy response. These concerns are especially important in trafficking research, where target populations may be heterogeneous and often difficult to reach through a single or traditional sampling strategy. This study examined whether the Prevalence Reduction Innovation Forum (PRIF) indicators developed by a group of global experts can support comparable trafficking severity scores across two countries and two sampling contexts, and whether item-level differences alter how trafficking severity should be interpreted in population-based social welfare research. The broader goal was to assess whether trafficking measurement can move beyond binary classification toward more interpretable, human-centered, policy-relevant severity scoring.

Methods:
Data came from 7,344 youth and young adults ages 18 to 37 in Malawi and Zambia, including both household samples and higher-risk link-tracing samples. Analyses focused on 32 binary PRIF indicators spanning the following domains: recruitment, employment practices, personal life and property, degrading conditions, freedom of movement, debt or dependency, and violence. We evaluated whether the item pool reflected a common trafficking severity continuum, then examined whether items operated similarly across countries and sampling frames. We also assessed whether the resulting severity scores aligned with a direct forced-labor question and with existing PRIF threshold-based summary variables. The analytic goal was to determine whether scores derived from these indicators could support meaningful comparisons across groups relevant to social work research and intervention planning.

Results:
Findings showed that the PRIF indicators can be used to capture differences in trafficking severity, beyond binary classifications. Although not all indicators functioned equivalently across countries and sampling contexts, 7 items remained stable across both comparisons and supported anchor-based cross-group comparison. In the higher-risk link-tracing sample, respondents with higher PRIF severity scores were substantially more likely to report direct forced labor (OR = 5.00). The PRIF severity score also closely matched existing PRIF threshold-based summaries (AUCs = .955-.977), suggesting it is consistent with earlier approaches while providing a more detailed picture of trafficking severity. Overall, the findings indicate that some indicators are context-sensitive, but that the broader PRIF measure can still support meaningful severity scoring across groups.

Conclusions and Implications:
These findings suggest that trafficking measurement in social work research can move beyond single indicators and binary thresholds toward more nuanced, person-centered severity scoring. At the same time, cross-country and cross-sample comparisons should not assume perfect equivalence. For social work, this has direct implications for identifying high-risk populations, comparing vulnerability across settings, monitoring program impact, and improving intervention targeting. More broadly, the study shows that stronger trafficking measurement can support better decisions for individuals, service systems, communities, and policy efforts aimed at prevention, response, and resource allocation.