3.2.1 Antecedents belonging to contingency theory
To detect the rate of unionization, we collected data on the percentage of employees organized in unions from the 10-K filings (the same operationalization is used by the Bureau of Labor Statistics; Blanchflower 2007; Faber 2005). Radical workforce changes were captured by the difference between the minuend number of employees in t − 2 and the subtrahend number of employees in t − 1 (1-year time lag included) (Kivimäki et al. 2001). The knowledge intensity of a firm was captured by the ratio of research and development (R&D) costs to sales or R&D intensity. This measure was introduced by the OECD in the mid-1980s, and has since then been widely used as an indicator of knowledge intensity (see OECD 2001, and as examples, Coff2003; Leydesdorff et al. 2006). These data were collected from S&P’s Compustat. The values for the antecedents “Outsider” and “New CEO” were collected from the 10-K filings. “Outsider CEO” was captured as a binary variable, with a value of 1 if the CEO had joined the firm no more than 2 years prior to becoming a CEO, and 0 otherwise (cf. Cannella and Lubatkin, 1993; Hambrick and Cannella 2004; Zhang 2006). “New CEO” was captured as a binary variable with a value of 1 if the CEO was in their first 3 years in the position, and 0 otherwise (cf. Hambrick and Cannella 2004).