Date of Award
5-2026
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Industrial Engineering
Committee Chair/Advisor
Dr. Thomas Sharkey
Committee Member
Dr. Emily Tucker
Committee Member
Dr. Yongija Song
Abstract
Human trafficking remains one of the most pervasive and complex human rights violations occurring within the United States, and South Carolina is no exception. From the 2023 South Carolina Human Trafficking Task Force annual report, there were 357 identified trafficking cases in the state, with 498 victims involved, providing the baseline survivor count that this thesis scales upward to account for undetected prevalence. From the Safe House Project, "Without proper aftercare, 80% of trafficking victims will be re-victimized," which is the direct motivation for this research's focus: sustained services must be prioritized rather than just emergency response alone. The 2014 State Plan identified a "lack of sufficient funding for, access to, and availability of resources for groups that provide services to victims of human trafficking," a gap that persists over a decade later, anchoring the problem statement of this thesis.
The primary question this research aims to answer is: how can a mathematical model be built, and data be gathered to populate it, so that South Carolina can optimize resource allocation across service organizations to minimize unmet demand while maintaining equitable access across counties? The objectives of this research are: (1) characterize baseline service capacity, (2) estimate county-level survivor demand, (3) develop and solve an optimization model for service expansion and resource allocation to better meet demand; and (4) conduct sensitivity analyses across budget, equity weighing, and equity measurement.
This research has developed the first comprehensive state-level optimization model integrating location decisions, capacity expansion, and multi-service allocation for varying survivor profiles. A novel tri-metric approach to measuring unmet demand was created (profile-driven, service-driven, individual-driven). The model incorporates 20 distinct survivor profiles, 12 service categories, and 27 provider organizations across South Carolina's 46 counties, utilizing real-world cost and capacity data from organizations including Jasmine Road and The Formation Project. The objective function balances equity (meeting complete needs for maximum survivors) against efficiency (maximizing total services delivered), controlled by parameter λ.
A series of 13 experiments were run across 3 sensitivity analyses using Python with Gurobi. Before optimization, baseline metrics show 9,944 unmet service units (out of 32,320 total demand), only 22 fully met profiles (out of 920), and 69.3% total coverage. These baseline results indicate that nearly one third of survivor service needs go unmet, and survivors are rarely getting all the services they need.
Lambda sensitivity analysis reveals that equity-focused approaches (λ = 0.01–0.5) deliver 2.5% more fully met profiles while sacrificing less than 1% overall coverage compared to pure efficiency maximization. The recommendation is that a lambda value between 0.3 and 0.5 produces statistically better outcomes while providing essentially the same coverage. Budget sensitivity analysis shows steep improvement from $150,000 to $250,000 followed by flattening above $250,000, indicating an ideal budget threshold where marginal returns begin declining. The cost to complete one more survivor's service coverage increases by 77% above $250,000, indicating rapidly diminishing equity gains. Profile-driven allocation mode achieves the highest fully met count and is recommended for the State Task Force.
This work bridges a critical gap in existing literature. While substantial operations research work addresses trafficking network disruption, far less addresses survivor support infrastructure. The research demonstrates that publicly available Task Force reports provide sufficient information for strategic planning, and the methodology requires only standard Task Force annual reports (available in 47 states), making this framework immediately adaptable to several surrounding states.
Recommended Citation
Redmon, Taylor A., "Helping to Optimize Survivor Support Networks: A Decision-Support Model for Service Capacity and Resource Allocation in South Carolina" (2026). All Theses. 4714.
https://open.clemson.edu/all_theses/4714