In addition, we take a further step to examine the effect of non-media award-winning CEO and firms' innovation success. Using a set of non-media awards helps improve the precision of our test for the roles of award-winning CEOs in firms' innovation performance. The effect of non-media award on corporate innovation is weakened in the two and three years after the award year.
These advantages make investments in innovation more accessible and ultimately increase the innovative activities of companies with non-media award-winning CEOs. First, compared to media-based awards, non-media award winners are not always the center of media attention. In this paper, we carefully consider the impact of CEOs' non-media awards on corporate innovation consistent with the media-based awards.
We expect that companies led by non-media-awarded CEOs will achieve better innovation success. Companies led by non-media award-winning CEOs subsequently obtain more innovation output than comparable companies led by non-winning CEOs.” The full list of media and non-media awards can be found in Appendices A and B respectively.
Methodology
Therefore, in our baseline analysis we include several control variables for CEO characteristics such as CEO's age (CEO_Age), Tenure (CEO_Tenure), and gender (Female). CEO characteristics.18 Specifically, for all firms in our sample, we set the binary dependent variable to one if the firm's CEO won the award in the current year and zero otherwise. We include dummies for year and industry to control for variation over time and industry.19 In this setting, we assume that the criteria for selecting award winners for media awards and non-media awards are the same.20.
Nevertheless, including non-media awards in our analysis is helpful in disentangling the effect of media coverage on the relationship between CEO awards and corporate innovation. In our analysis, we examine the effect of winning CEO awards on firms' innovation output over the one-year (k=1), two-year (k=2) and three-year (k=3) periods following the award ceremony. Media_Award, a dummy variable, equals one if the company's CEO wins media awards in the current year, and zero if the CEO is the predicted winner.
Equation (2) is similar to equation (1), except that we replace Media_Award with the dummy variable Non-Media_Award, which takes the value of one if the CEO wins non-media awards in the current year and zero if the CEO is a winner anticipated.
Results
Non-media-awarded CEOs are older, have longer tenures and are more likely to be women. There are a number of non-media awards given specifically to female CEOs, which may lead to the positive and significant coefficient for the Woman variable.22 Furthermore, non-media award-winning CEOs are significantly differentiated from non-winners in terms of educational backgrounds. , demographic factors and experience. The Business Equipment industry group is significantly (at the 1% level) overrepresented among the non-media winners.
22 An example of a non-media award given only to women is the National Association Female Executives' Women of Excellence Avatar Award. Columns (1) and (2) of Table 3 report the results of the logit model for predicting media and non-media prices, respectively. It suggests that the award panels for non-media awards take into account other factors not reflected in previous results as criteria for selecting the winners.
Non-media winners are more likely to be in financial or technical fields, have military training, hold an individual patent, and are more likely to be born outside the US than their intended winners. The coefficients for innovation results, Patentt+1 and Citation t+1, in the year immediately following the year of award are positive and. statistically significant at the 1% level, suggesting that firms acquire more patents and citations in the one-year period after their CEOs win non-media awards. Recalling that in Table 2 the R&D spending of non-media winners and predicted winners is insignificantly different, the regression results show that after winning a non-media award, companies led by non-media award winners generate statistically higher corporate innovation output with relatively similar innovation inputs compared to companies run by announced winners.
In terms of economic importance, companies that are not media winners generate on average 0.53% more in number of patents and 0.38% more in number of citations compared to predicted winners in the first year after the award. The effect of winning a non-media award on corporate innovation is persistent in the second and third years following the award year. Two years after the year of award, companies that are not media winners are awarded 0.46% more patents and their patents are cited 0.34% more than predicted winners.
Therefore, the effect of winning non-media awards can be gradually transferred to innovation success. These findings hold for both media and non-media award samples and are robust to periods of one year, two years, and three years after the award year.
Robustness checks
In Table 5 , we find that with respect to media allocation, controlling for additional CEO characteristics does not change the results of the baseline regressions. In Panel A, where we report the results for a year after the award year, the coefficients on Non-media_Award are all statistically significant across all model specifications. Thus, the effect of winning a non-media award on corporate innovation is not driven by the above CEO characteristics.
Regarding the two-year and three-year periods following the award year (reported in panels B and C), coefficients for Non-media_award remain positive and significant after controlling for CEO education and demographic background. The coefficients on Non-media_Award are statistically significant at the percentage level for all of the two years following the year of award and then become weaker in the third year following the year of award. Second, we exclude the last two years from the sample to ensure that our results are not subject to potential truncation bias.
We find that the negative effect of winning Media award becomes insignificant when different award samples are considered. Results of this robustness test, reported in Row (6) of Panel A and Row (6) of Panel B, show that the effect of winning CEO awards (either Media or Non-Media) on innovation activity is independent of the effect of stock liquidity on innovation. According to Row (9) of Panel A, the coefficients for Media_award become insignificant after controlling for GIndex.
With regard to non-media awards, the inclusion of GIndex reduces the persistence of the effect of winning non-media awards on firm innovation. Specifically, as suggested by the results in row (9) of panel B, the non-media winning firms only have better innovation output than their predicted winners in the first year after the award year. The effect disappears in the second and third years following the award year after controlling for GIndex.
In contrast, the effect of winning non-media awards on corporate innovation in the first year after the award year is highly significant and robust. The long-run effect of non-media award winnings, however, becomes weaker in some of the robustness checks.
Conclusion
The effect of the non-media award on corporate innovation weakens in two and three years after the award year. Our finding that firms led by non-media award winners seem to engage in more corporate innovation outcomes is consistent with the view that the non-media award is a less biased (and therefore better) indicator of personal competence and managerial ability. Non-media award winners are less likely to be in the media spotlight; therefore they do not suffer from the issue of "famous burden".
Our results point to a fruitful avenue for future research to further examine the informational content of non-media CEO awards (currently neglected in the literature) in other corporate contexts. A dummy equal to one if the CEO receives at least one media award in a given year, and zero otherwise. A dummy equal to one if the CEO wins at least one non-media award in a given year, and zero otherwise.
A dummy that takes the value of one if the CEO has an MBA degree and zero otherwise. A dummy equal to one if the CEO attended an Ivy-League institution and zero otherwise. Data on CEO's Media and Non-media Awards is hand-compiled from CEO biographies in the Marquis Who's Who database.
CEO media and non-media awards data is hand-collected from CEO biographies in the Marquis Who's Who database. Columns (1) and (2) of the table report results for logit models predicting media and non-media award winners. The table reports regression results for the sample that includes non-media winners and their predicted winners.
Independent variables include Non-media_award (a dummy variable equal to one if the CEO wins at least one non-media award in year t and nothing else) and a set of control variables for CEO characteristics and firm characteristics similar to those in Table 4 (not shown due to brevity). Panels A and B show the regression estimates of Media_award and Non-media_award, respectively, in a regression model with patent, citation, and R&D as dependent variables. Non-media_award is a dummy variable equal to one if the CEO wins at least one non-media award in year t and zero otherwise.
In row (3) of panel B, awards from Ernst & Young are excluded from non-media award lists.
Appendix A. List of CEO Media Awards
Appendix B. List of CEO Non-Media awards