Recursive Inspection Games

TL;DR

This paper models a sequential inspection game with limited inspections and varying rewards, providing explicit solutions and extending to non-zero-sum and leadership scenarios, advancing strategic inspection theory.

Contribution

It introduces a recursive framework for inspection games with variable rewards and solves the equilibrium explicitly, also exploring non-zero-sum and inspector leadership extensions.

Findings

Explicit recursive solution for equilibrium payoffs.

Strategy independence from number of violations under certain conditions.

Extension to non-zero-sum and leadership scenarios.

Abstract

We consider a sequential inspection game where an inspector uses a limited number of inspections over a larger number of time periods to detect a violation (an illegal act) of an inspectee. Compared with earlier models, we allow varying rewards to the inspectee for successful violations. As one possible example, the most valuable reward may be the completion of a sequence of thefts of nuclear material needed to build a nuclear bomb. The inspectee can observe the inspector, but the inspector can only determine if a violation happens during a stage where he inspects, which terminates the game; otherwise the game continues. Under reasonable assumptions for the payoffs, the inspector's strategy is independent of the number of successful violations. This allows to apply a recursive description of the game, even though this normally assumes fully informed players after each stage. The resulting recursive equation in three variables for the equilibrium payoff of the game, which generalizes several other known equations of this kind, is solved explicitly in terms of sums of binomial coefficients. We also extend this approach to non-zero-sum games and, similar to Maschler (1966), "inspector leadership" where the inspector commits to (the same) randomized inspection schedule, but the inspectee acts legally (rather than mixes as in the simultaneous game) as long as inspections remain.