Organizational slack and performance: the impact of outliers

Abstract

Extending initial theory development and empirical studies conducted in the early 1980s, the investigation of the relationship between slack resources and the performance of the firm has experienced renewed attention. Over the past 25 years enough empirical research has been conducted that researchers have begun to question the frequently found ambiguous results. We assess the impact statistical outliers may have on the relationship between organizational slack and firm performance trying to determine whether the positive, curvilinear, or ambiguous results that have been reported are impacted by the presence of outliers. We found that the measures of organizational slack were highly variable due to five general factors. When the outliers, extreme values, were statistically identified and removed, the relationship between organizational slack (Available Slack) and performance (ROA) became non-linear and consistent over years. Implications suggest that future research should consider the potential impact that non-normal distributed data could have on the validity of findings, particularly when employing data from archival sources. Suggestions for future research in the slack-performance relationship are offered.