To test the hypotheses, standard ordinary least squares regression was applied. For all models, the variance inflation factor was calculated to check for potential multicollinearity.The highest value across all models and variables refers to industry SIC 35 in model 5 andamounts to 2.56. This value is well within an acceptable range (Hair et al., 2006). In addition, the residuals in all models have been checked for normal distribution by applying the Komolgorov–Smirnov test with Lilliefors’ modification. According to this test, the residuals in all models are normally distributed at least at the level of p < .1 (Hair et al., 2006). To cross-validate the results of the OLS regression analyses, structural equation modeling was employed. The structural equation models show no significant changes in the findings compared with the results of the OLS regression analyses. Accordingly, for reasons of parsimony, the paper presents only the results of the OLS analyses in detail.