Table 5. Estimation result (dependent variable: annual R&D investment per stock)

Specification GMM-SYS Robust check
Fixed effect OLS Pooled OLS
Spec. 1 Spec. 2 Spec. 3 Spec. 4. Spec. 5 Spec. 6.
λ1Ri,t1Ci,t2 –0.001*** (0.000) –0.001*** (0.000) –0.001*** (0.000) –0.001*** (0.000) –0.001*** (0.000) –0.001*** (0.000)
γ2Δyi,t 0.083*** (0.028) 0.079*** (0.027) 0.079** (0.030) 0.076*** (0.030) 0.110*** (0.026) 0.105*** (0.025)
γ3Δyi,t-1 0.094*** (0.025) 0.100*** (0.025) 0.103*** (0.025) 0.107*** (0.025) 0.089*** (0.025) 0.098*** (0.024)
γ4 (ci,t-1yi,t-2 –0.198*** (0.029) –0.196*** (0.029) –0.276*** (0.043) –0.273*** (0.043) –0.175*** (0.029) –0.173*** (0.029)
γ5yi,t-2 –0.118*** (0.021) –0.111*** (0.021) –0.168*** (0.030) –0.161*** (0.030) –0.101*** (0.019) –0.093*** (0.019)
ϕ1Zombie statusi,tCi,t1 5,158.9*** (468.97) 5,084.6** (468.56) 5,613.1*** (462.81) 5,503.0*** (442.04) 5,649.9** (460.71) 5,542.9** (433.46)
ϕ2Zombiei,t×ILZSiCi,t1 - 149,409.3** (60,078.25) - 139,443.4** (65,582.93) - 155,759.3** (64,433.88)
Constant 0.000*** (0.000) 1.832*** (0.311) 2.492*** (0.427) 2.366*** (0.431) 1.728*** (0.317) 1.576*** (0.321)
Firm fixed Yes Yes Yes Yes No No
Tech. field fixed1) - - - - Yes Yes
Year fixed Yes Yes Yes Yes Yes Yes
Observations 2,981 2,981 2,981 2,981 2,3372) 2,337
R-squared - - 0.114 0.118 0.130 0.139
Number of firms 704 704 704 704 - -
Number of instruments 21 22 - - - -
F-statistic 264.28 190.32 1,752.65 3,728.93 84.10 201.35
P (F-statistic) 0.000 0.000 0.000 0.000 0.000 0.0000
AR test (p-val) 1st order 0.109 0.106 - - - -
      2nd order 0.558 0.599 - - - -
Hansen J test (p-val) 0.087 0.091 - - - -
are statistical significance at 1%,
5%,
10% respectively; standard errors are also reported in parentheses; p-values are reported for AR test and Hansen J test.
Tech. field includes the followings: 1 (machine), 2 (electronics), 3 (information technology), 4 (chemical), 5 (bio), 6 (energy), 7 (knowledge service), 8 (ceramics).
In the pooled OLS estimation, the technology field is newly added unlike GMM-SYS and Fixed effect OLS estimation. Some missing data in technology field make difference in the number of total observations. R&D, research and development; GMM-SYS, Generalized Method of Moments System; OLS, ordinary least squares; ILZS, industry with lower zombie share; AR, auto-regressive.