Based on underlying literature we develop an ex ante control framework which suggests that a lean manufacturing
strategic initiative directly affects organizational structure, accounting practices, and control components. In this ex
ante model organizational structure is the only intervening variable between the lean manufacturing strategic initiative,
and accounting and control components. We expect the relations in the model to be best characterized by a bivariate,
Cartesian fit. We find that the data at the field site reveal both accounting practices and organizational structure as two
important intervening variables in the relation between a lean manufacturing strategic initiative and management control.
Although we see evidence that the lean manufacturing initiative directly influences EBS’ selection of and reliance
on control components, we also see evidence that both the organizational structure and accounting practices influence
the design and reliance EBS places on the control components. EBS shifted its control from an emphasis on variance
analysis feeding a performance measurement system intended to direct effort, to a control system that now includes a
portfolio of controls that emphasizes output, behavioral, and social controls. Unlike the ex ante control framework, we
also observed many unidirectional and bidirectional relations within the portfolio of control components. For example,
the performance measurement system (i.e., an output control) tracks cross-training and thus it motivates employees to
pursue additional training (i.e., a social control). Another example is that the performance measurement system and
employee empowerment (i.e., a social control) influence each other. The employees are empowered to collect, produce,
record, and report their own performance measurement information, which also empowers the employees by providing
them with the information needed to facilitate effective decision-making. We construct a theoretical framework based
on our observations and the evidence gathered at EBS in which we conclude that the relations in the revised framework
are better characterized by a Configuration fit. We contrast our theoretical framework with prior research and discuss
implications.