presents a summary of our sample. We assessed the possibility of non-response bias by following De Luca and Atuahene-Gima [45]. Specifically, we compared a sample of 50 matched firms with a sample of unmatched firms for which we had data on firm age and number of employees. Analyses of variance indicated no significant differences between the two groups in terms of firm age (F = .50) and the number of employees (F = .42).