This project investigates a significant feature of the socioenvironmental context of illicit drug use. Directly or indirectly, all users engage the illicit drug economy. Understanding the market roles of producers, consumers, and distributors is critical for prevention, treatment and law enforcement. However, drug markets are dynamic;they continuously adapt to internal and external forces. No theory (or methodology) exists to describe such dynamics. This five-year study develops and systematically field-tests agent-based modeling (ABM) methods to represent consumer, dealer, and health behaviors associated with the methamphetamine market in Cuyahoga and Summit Counties, OH. This approach will provide researchers with an entirely new perspective for observing outcomes of interactive behaviors, as well as the opportunity to perform experiments not possible in the real world. To program simulations, we will characterize roles, motives, behaviors, and interactions of market participants utilizing ethnographic methods (specifically, decision-tree modeling techniques). To validate and extend these data, we will collect quantitative measures of participants'daily behaviors, drug consumption, costs, and interactions using Ecological Momentary Assessment (EMA). We will implement these techniques through a three-wave panel study (N=204);data from the panel study will characterize the sample, inform simulation parameters, and help validate simulation output. Our simulations will subsequently incorporate both descriptive ethnographic data and quantitative longitudinal data so that the behaviors of """"""""agents"""""""" (simulated people interacting within the virtual market) will accurately reflect the realities of the real-world drug market. The simulation-based computer laboratory produced by this project will offer policymakers and researchers a tool for: (1) dynamically representing illicit drug market operations, (2) experimenting with agent behaviors and market parameters and conditions, and (3) evaluating """"""""what if"""""""" policy scenarios intended to influence outcomes. Applying Complexity theory by systematically combining the rich detail of ethnography with ABM's power to aggregate socially complex behaviors, this project will have great practical utility.
The specific aims of this project are to: (1) Conduct ethnographic research on the methamphetamine market in Cuyahoga and Summit Counties, OH. (2) Enrich these data using Ecological Momentary Assessment (EMA), collecting self-report data on daily drug consumption, production, sales, decisions, strategies etc. (3) Inform simulation parameters using a panel survey of 204 active methamphetamine users. (4) Construct a computer lab of ABM simulations reproducing how the local methamphetamine market operates integrating both social (i.e., health) and economic behaviors. (5) Experiment with the ABM simulations to: understand how the market operates and functions;create and test policy-based intervention scenarios (e.g. enforcement, treatment, and outreach) intended to impact outcomes;and model risk behaviors (e.g., needle sharing, trading drugs for sex) influencing the spread of HIV.
Drug users risk behaviors are influenced by the markets in which they acquire drugs. However, these markets are dynamic and reactive which make understanding and modeling this interaction challenging. The objective of this research is to field test a transportable, theoretically informed, and inter-disciplinary research methodology using agent-based simulation to evaluate market dynamics and their impact on methamphetamine distribution and HIV-related risk behaviors.
|Bobashev, Georgiy; Holloway, John; Solano, Eric et al. (2017) A Control Theory Model of Smoking. Methods Rep RTI Press 2017:|
|Bobashev, Georgiy; Norton, Jacob; Wechsberg, Wendee et al. (2014) Are you HIV invincible? A probabilistic study of discordant couples in the context of HIV transmission. PLoS One 9:e94799|
|Heard, Daniel; Bobashev, Georgiy V; Morris, Robert J (2014) Reducing the complexity of an agent-based local heroin market model. PLoS One 9:e102263|