ENVIRONMENTAL PRESSURE, ORGANIZATIONAL BUFFERS AND ORGANIZATIONAL PERFORMANCE: A STRUCTURAL EQUATIONS MODEL (SLACK).

Abstract

This dissertation addresses questions concerning slack's nature and its relationships with the environment and performance. The research investigates which view of slack (the operations or behavioral approach) best predicts performance. It examines the relationship of environment and slack using both interaction and mediation models. The PIMS database was used for 610 assembly manufacturing firms. The results support both the behavioral and the operations perspectives. This combined view suggests that slack capacity is optimized to improve sales while being minimized to improve profits. Excess inventory is minimized to improve sales but optimized to improve average ROS. In all cases, excess cash is minimized. In all equations, the slack variables entered the equations as costs. These results also support the argument that slack interacts with the environment rather than being in a functional relationship with it. Interaction terms of the slack types and the environment were significant in predicting sales. A mediation model was also tested but had a poorer fit with the data. Slack was found to be a multi-dimensional concept. The slack variables did not all intercorrelate positively. The negative relationships suggest that management makes decisions as when to use each slack resource. The slack variables (when lagged) had significant effects on each other, but not on performance. This indicates that the time horizon for slack may be shorter than was investigated in this research. The research demonstrated that slack inventory and non-slack supply buffers were negatively related. The conditions under which the firm trades slack for other buffering mechanisms were not clear. Predicted positive relationships between size and slack were found except that excess capacity and size were negatively related. This suggests that larger firms were holding slack in ways that are more discretionary and less obvious to their control systems. What is not clear from this research are the conditions under which management will choose a specific type of slack. In one case (excess working capital), technology predicts the level of this variable. Additional research is suggested to determine how, when and where these decisions are made.