Numerical implementation code for 'Capturing the Enforcer: A Dynamic Principal-Agent-Client Model of Corruption and Institutional Persistence

This repository contains the Python implementation of the numerical model in "Capturing the Enforcer: A Dynamic Principal-Agent-Client Model of Corruption and Institutional Persistence." The model characterizes a discrete-time dynamic principal-agent-client framework in which institutional integrity is an endogenous state variable eroded by corruption and restored by institutional resilience.

The code solves the agent's dynamic optimization problem by value-function iteration on a discretized state-control grid, locates stationary institutional equilibria via Brent's method, and classifies their local stability. It reproduces Table 1 (baseline stationary equilibria) and Figures 2–7 of the paper, including the induced transition map, the optimal corruption, bribe, and reelection-probability schedules, simulated institutional trajectories, and the bifurcation analysis for institutional resilience, capture capacity, legal enforcement, social tolerance, and electoral incentives.

Running institutional_capture_model.py with no modifications reproduces all reported results under the baseline calibration (Appendix B, Table B1) and writes the corresponding tables and figures to a results/ directory.

Requirements: Python 3, NumPy, SciPy, pandas, Matplotlib.