Effect of TMT Diversity on innovation outcomes (baseline, alternatives, and checks)
| Outcome | Baseline β (SE) | Alt. Aggregation β (SE) | Robust SE β (HC1) | Leave-one-out range |
|---|---|---|---|---|
| Patents (ln(1 + patents)) | −0.39 (0.26) | −1.02 (0.72) | −0.39 (0.18) | [−0.53, −0.29] |
| Product releases | +1.89 (0.78) | +5.11 (2.15) | +1.89 (0.73) | [+1.34, +2.42] |
| Digital revenue share (%) | +2.01 (11.41) | +2.85 (31.20) | +2.01 (7.36) | [−1.23, +8.22] |
| Outcome | Baseline | Alt. Aggregation | Robust SE | Leave-one-out range |
|---|---|---|---|---|
| Patents (ln(1 + patents)) | −0.39 (0.26) | −1.02 (0.72) | −0.39 (0.18) | [−0.53, −0.29] |
| Product releases | +1.89 (0.78) | +5.11 (2.15) | +1.89 (0.73) | [+1.34, +2.42] |
| Digital revenue share (%) | +2.01 (11.41) | +2.85 (31.20) | +2.01 (7.36) | [−1.23, +8.22] |
Note(s): All models include controls for CDE, Integration, and year (with firm fixed effects throughout); continuous predictors are mean-centered. Complete robustness and sensitivity analyses—including alternative diversity aggregation, HC1 robust standard errors, and leave-one-out diagnostics—are reported in Appendix D (Table D1)