The university research environment has been undergoing profound change in recent decades. Aiming at international competitiveness and excellence, a variety of polices have been designed and implemented in many countries. However, evidence-based analysis of policy effects is scarce. This article develops methods for evaluating the effect of university research policy on university system research input-output dynamics. We assume stable dynamics between inputs and outputs and that effective policy change introduces external interventions and therefore structural changes into the system. Our proposed method involves three steps: modeling system dynamics, detecting structural change, and mapping policy change. Examining the case of the UK, we identified three structural changes in 1990, 1994, and 2001 respectively. The first change corresponds to the second round of the Research Selectivity Exercise and the latter two changes may be caused by the database effect. The British case is a provocative demonstration of this method, which could be further developed to provide evidence-based analysis for policy learning and a foundation for cross-case comparison.