Abstract
This longitudinal study investigates the integration of the controllability principle in the redesign of a performance measurement system (PMS). The PMS and the controllability principle are conceptualized as “epistemic objects” (open, question-generating, and complex objects) which are associated with varying stakeholder desires. The paper makes two contributions. First, it demonstrates how a PMS can oscillate between different interpretations of the controllability principle (narrow versus broad), resulting in different controllability boundaries (tight versus loose). Second, the paper theorizes how multiple epistemic objects are formed into combination structures (a combination of epistemic objects that are subsequently codeveloped). Three sequentially developed combination structures of the controllability principle and the PMS, along with their associated effects, are reported. The paper demonstrates how combination structures can, ironically, become antithetical to the desires that set change in motion, and in this case, render the PMS incapable of measuring performance on complex environmental and societal challenges.