CEO research orientation - Appendices - FINAL.pdf

Full text

Appendix
CEO Research Orientation, Organizational Context, and Innovation in the
Pharmaceutical Industry

Appendix A: Analyses using granted patents
Table A: Results of when CEO research orientation affects innovation
Dependent variable
Innovation outcomes
1
2
3
4
5
6
CEO research orientation
0.16***
0.05
0.10
0.26***
0.06
(0.05)
(0.08)
(0.06)
(0.06)
(0.09)
CEO RO*CEO duality
0.14
0.15
(0.09)
(0.09)
CEO RO*Financial slack
0.01**
0.01**
(0.00)
(0.00)
CEO RO*Firm age
-0.00*
-0.00
(0.00)
(0.00)
Firm size
0.42***
0.45***
0.45***
0.44***
0.45***
0.44***
(0.06)
(0.06)
(0.06)
(0.06)
(0.06)
(0.06)
Firm age
-0.01
-0.01
-0.01
-0.01
-0.01
-0.01
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
Financial slack
0.01*
0.01
0.01
-0.02
0.01
-0.02*
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
Financial performance
-0.19
-0.20
-0.21
-0.16
-0.18
-0.16
(0.11)
(0.11)
(0.11)
(0.11)
(0.11)
(0.11)
Institutional ownership
-0.00
-0.00
-0.00
-0.00
-0.00
-0.00
(0.00)
(0.00)
(0.00)
(0.00)
(0.00)
(0.00)
CEO tenure
0.00
0.00
0.01
-0.00
0.00
0.01
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
Founder
0.14
-0.14
-0.16
-0.18
-0.21
-0.25
(0.13)
(0.15)
(0.15)
(0.15)
(0.15)
(0.16)
CEO duality
0.06
-0.03
-0.19
-0.10
-0.04
-0.29*
(0.11)
(0.11)
(0.13)
(0.10)
(0.11)
(0.12)
CEO ownership
0.01
0.01
0.01
0.01
0.01
0.00
(0.01)
(0.01)
(0.01)
(0.02)
(0.01)
(0.01)
Insider
-0.04
-0.10
-0.08
-0.16
-0.13
-0.16
(0.10)
(0.10)
(0.10)
(0.10)
(0.10)
(0.11)
Board independence
-0.23
-0.21
-0.20
-0.32
-0.25
-0.33
(0.40)
(0.38)
(0.37)
(0.39)
(0.37)
(0.37)
Presample patent stock
0.00
0.00
0.00
0.00
0.00
0.00
(0.00)
(0.00)
(0.00)
(0.00)
(0.00)
(0.00)
Presample dummy
-2.27***
-2.30***
-2.37***
-2.42***
-2.39***
-2.56***
(0.49)
(0.43)
(0.44)
(0.48)
(0.42)
(0.47)
Lagged dependent variable (patents)
0.01***
0.01***
0.01***
0.01***
0.01***
0.01***
(0.00)
(0.00)
(0.00)
(0.00)
(0.00)
(0.00)
Constant
-1.80**
-2.07***
-2.04***
-1.69***
-2.03***
-1.65**
(0.56)
(0.52)
(0.52)
(0.51)
(0.51)
(0.52)
Observations
711
711
711
711
711
711
QIC
5178.4
5118.0
5105.9
5116.3
5119.1
5103.3
Wald chi-square
616.7***
671.9***
670.4***
854.5***
741.8***
872.2***
Note: Robust standard errors in parentheses. All models include SIC and time dummies. Significance levels of
two-tailed tests: * p <.10; ** p<.05; *** p<.01.

Table B: Results of how CEO research orientation affects innovation
Dependent variable
R&D intensity
Innovation outcomes
7
8
9
10
11
12
CEO research orientation
8.62*
0.16**
0.07
0.21***
0.12
(4.27)
(0.05)
(0.09)
(0.05)
(0.09)
R&D intensity
0.15
0.15
(0.09)
(0.09)
CEO RO*CEO duality
0.01**
0.01**
(0.00)
(0.00)
CEO RO*Financial slack
-0.00*
-0.00
(0.00)
(0.00)
CEO RO*Firm age
0.00**
0.00**
0.00***
0.00***
(0.00)
(0.00)
(0.00)
(0.00)
R&D intensity*CEO RO
-0.00*
-0.00*
(0.00)
(0.00)
Firm size
-23.53***
-22.07***
0.49***
0.48***
0.49***
0.48***
(5.64)
(5.42)
(0.06)
(0.06)
(0.06)
(0.06)
Firm age
0.83
0.83
-0.01
-0.01
-0.01
-0.01
(0.54)
(0.53)
(0.01)
(0.01)
(0.01)
(0.01)
Financial slack
1.64*
1.45*
0.01
-0.02
0.01
-0.02
(0.68)
(0.71)
(0.01)
(0.01)
(0.01)
(0.01)
Financial performance
29.72
30.56
-0.12
-0.09
-0.11
-0.08
(16.85)
(16.86)
(0.11)
(0.12)
(0.11)
(0.12)
Institutional ownership
0.44*
0.38*
-0.00
-0.00
-0.00
-0.00
(0.19)
(0.19)
(0.00)
(0.00)
(0.00)
(0.00)
CEO tenure
-0.52
-0.58
0.00
0.01
0.00
0.01
(0.86)
(0.81)
(0.01)
(0.01)
(0.01)
(0.01)
CEO founder
7.51
-3.55
-0.14
-0.26
-0.10
-0.22
(11.03)
(10.25)
(0.15)
(0.16)
(0.13)
(0.15)
CEO duality
-9.02
-10.29
-0.04
-0.30*
-0.05
-0.30**
(9.66)
(9.53)
(0.11)
(0.12)
(0.11)
(0.11)
CEO ownership
-1.54
-1.55
0.01
0.00
0.01
0.00
(0.94)
(0.97)
(0.01)
(0.01)
(0.01)
(0.01)
Insider
6.88
2.05
-0.12
-0.18
-0.08
-0.15
(12.96)
(12.33)
(0.10)
(0.12)
(0.10)
(0.11)
Board independence
-30.43
-29.32
-0.25
-0.37
-0.27
-0.38
(30.69)
(31.26)
(0.39)
(0.37)
(0.38)
(0.37)
Presample patent stock
0.02
0.02
0.00
0.00
0.00
0.00
(0.01)
(0.01)
(0.00)
(0.00)
(0.00)
(0.00)
Presample dummy
-41.55
-43.61*
-2.22***
-2.48***
-2.13***
-2.39***
(23.50)
(19.04)
(0.43)
(0.45)
(0.42)
(0.44)
Lagged dependent variable (R&D)
0.43***
0.43***
(0.05)
(0.05)
Lagged dependent variable (patents)
0.01***
0.01***
0.01***
0.01***
(0.00)
(0.00)
(0.00)
(0.00)
Constant
133.59***
126.83***
-2.29***
-1.87***
-2.46***
-2.01***
(39.10)
(36.57)
(0.53)
(0.53)
(0.52)
(0.52)
Observations
711
711
711
711
711
711
QIC
5277543.5
5118918.3
4461.9
4466.0
4457.0
4461.7
Wald chi-square
1698***
2119***
704.0***
998.1***
714.4***
1060***
Note: Robust standard errors in parentheses. All models include SIC and time dummies. Significance levels of two-
tailed tests: * p <.10; ** p<.05; *** p<.01.

Appendix B. Reliability and factor analysis of CEO research orientation
Cronbach’s alpha applied to the four items of the research orientation index equals .87, which
is well above the threshold of .70 recommended for evaluating the reliability of new constructs
(Nunnally, 1978). The tetrachoric correlations among the indicators were high and all positive,
ranging from .62 to .94 (p < .01). We used tetrachoric correlations because correlations among
binary variables are much lower than the correlations among continuous variables, which
makes the use of Pearson correlations problematic. To further confirm the coherence among
indicators, we conducted an exploratory factor analysis, again using tetrachoric correlations.
With a principal component factoring procedure, all four indicators loaded on a single factor
(with loadings ranging from .87 to .99) that had an eigenvalue of 3.48, explaining 87.1 of the
variance.
Table C: Tetrachoric Correlations of CEO Research Orientation Items
1
2
3
1
PhD engineering/science
2
Patent holder
0.78
3
R&D experience
0.94
0.86
4
Academic experience
0.82
0.62
0.92
Table D: Factor Loading of Exploratory Factor Analysis
Variable
Factor 1
PhD engineering/science
0.953
Patent holder
0.872
R&D experience
0.999
Academic experience
0.904

Appendix C: Endogeneity control
Following research in pharmaceutical innovation and upper echelons research (Gerstner et al.,
2013), we explored a broad set of possible predictors of CEO research orientation to address
potential concerns that result from the possibility that research-oriented CEOs might be
attracted by and selected in contexts that match their orientation. We did so by regressing the
109 CEOs against a set of carefully selected variables that potentially drive a CEO selection
effect. In various regression analyses, we included (combinations of) the following variables
measured at the year of appointment and/or at the year prior to appointment: CEO insider, CEO
age, CEO appointment year, calendar year of appointment, the number of years since the
establishment of the pharmaceutical industry (around 1972/1976), firm age, firm patent
propensity, firm R&D investment stock, firm patent stock, firm R&D intensity, industry age,
industry R&D intensity, and SIC codes.
Only firm age prior to appointment and CEO insider significantly predicted CEO research
orientation with an overall model R-squared of .12 (F = 2.35, p = .036). As an alternative to
OLS regression, we also ran an ordered probit regression predicting the research orientation
score of the CEO because 65 CEOs had a research orientation score of zero. The overall model
reported a McFadden’s pseudo R-squared of .06 (LR chi2 = 15.50, p = .030) and the findings
were the same as those reported in the OLS regression here. We then used the regression
coefficients for these variables to calculate each CEO’s predicted research orientation score
and include that value as an “endogeneity control” in all analyses. However, since only two
variables predicted CEO research orientation, We decided that the quality of the constructed
endogeneity control was too low and therefore included firm age and CEO insider as control
variables in all analyses.