Acknowledgements
I would like to thank Ali Soomro, Arun Jacob and Margarita Yakovleva for help with the data collection and analysis. I would like to thank Enrique Schroth for providing access to the innovation data and also for helpful discussions. The author thanks Sriya Iyer and participants of the Management Science and Judge-Us seminar series at the Cambridge Judge Business School, University of Cambridge, and the Editor-in-Chief for comments on earlier versions of this article.
Notes
†Zingales (Citation2000) argues that retaining employees with expertise and pursuing investment opportunities is one of the most challenging corporate governance issues for new economy firms that rely on knowledge and managerial capital.
†It could be argued that the new businesses that are less ideas-driven are effectively relationships-driven. Relationships-driven implies that the new business being set up by the former employee is primarily because of customer relationships taken from the previous employer.
†In order to verify the inter-rater reliability, we conducted the Proportional Reduction in Loss (PRL) reliability measure analysis (Rust and Cooil Citation1994) on our ratings of eight experts. The proportion of inter-rater agreement corresponded to 0.8, which is above the acceptable level of agreement between the experts on their ratings of the survey question (Nunnally Citation1978).
†Bonus payments were not reported by the banks for all the years of our sample. Therefore, we operationalized the measure by counting the number of articles in The Wall Street Journal and The Financial Times that mentioned the word ‘bonus’ in relation to a particular bank for the relevant years. The article count measure correlates significantly with the sub-sample of banks that did report their bonus payments and average compensation per employee for the period of the study, which provides support that our operationalization is a reasonable proxy measure for bonus payments. The credit ratings measure is based on ratings of the banks provided by Standard and Poor's (S&P).
†There is no readily available method for testing for endogeneity due to unobserved heterogeneity using the Heckman sample selection method in ordered probit models. Therefore, we used a probit specification to test for endogeneity using the Heckman procedure, where we coded scores of 2 or more for the degree of ideas-driven innovation as 1 and as 0 otherwise.