Industry segment effects and firm effects on firm performance in single industry firms

The purpose of the paper is to identify the sources of variation in firm performance. This is one of the cornerstones of strategy research, i.e. the relative importance of industry and firm level effects on firm performance. Multilevel analysis is well suited to analyze variance in performance when the data are hierarchically structured (industry segments consist of firms, firms operate within the context of industry segments). The Belgian industry studied is a service industry that consists of about 25 electrical wholesalers. Data were collected from 20 firms during the period 1998-2003 from responses to a questionnaire sent to all the firms in the market. The sample in the data set covers more than 95 percent of the market (in sales), as the missing firms were just fringe competitors. The results show that firm effects explain most of the variance in four performance variables. That bears out the importance of each firm having its own specific, idiosyncratic resources and competences. The explanatory power of firm effects varies by about 30 to 40 percent while the intra-industry effect explains around 10 percent of the variance. Even though firm effects are dominant, intra-industry effects explain a significant portion of the variance in firm level performance. The firm effect is smaller than in previous studies. The firm effect varies across the performance measures: firm effects are higher for returns on assets than for profit margins. The industry segment effect (or intra-industry effect) is more independent of the dependent variable. The industry segment effect is in line with previous studies on the strategic group effect. Top managers should carefully choose and monitor the intra-industry domain they are in.