Statistical Agent Based Modelization of the Phenomenon of Drug Abuse

TL;DR

This paper presents a statistical agent-based model to simulate drug abuse dynamics, considering individual heterogeneity and social factors, aiming to inform effective policy strategies.

Contribution

It introduces a novel agent-based modeling approach that incorporates heterogeneity and real-world data for understanding drug abuse evolution.

Findings

Rare personal events significantly influence drug initiation.

System perturbation analysis reveals key factors in drug abuse dynamics.

Model aligns well with available empirical data.

Abstract

We introduce a statistical agent based model to describe the phenomenon of drug abuse and its dynamical evolution at the individual and global level. The agents are heterogeneous with respect to their intrinsic inclination to drugs, to their budget attitude and social environment. The various levels of drug use were inspired by the professional description of the phenomenon and this permits a direct comparison with all available data. We show that certain elements have a great importance to start the use of drugs, for example the rare events in the personal experiences which permit to overcame the barrier of drug use occasionally. The analysis of how the system reacts to perturbations is very important to understand its key elements and it provides strategies for effective policy making. The present model represents the first step of a realistic description of this phenomenon and can be easily generalized in various directions.