Skip to main content

The Official Journal of the Pan-Pacific Association of Input-Output Studies (PAPAIOS)

Table 5 System-GMM estimation, with income level dummy—dependent variable: LN EMPSHARE

From: Does innovative capacity affect the deindustrialization process? A panel data analysis

Regressors

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

Model 7

Model 8

Model 9

Model 10

BASELINE

R&D

RESEARCHERS

TECHNICIANS

ARTICLE

PATENTS

TRADEMARK

INCENG

INDOUT

INES

LN EMPSHARE(-1)

0.869828***

0.678002***

0.856861***

0.721469***

0.778751***

0.810980***

0.844851***

0.804786***

0.705562***

0.633975***

 

(0.013829)

(0.015378)

(0.033870)

(0.024906)

(0.016856)

(0.018288)

(0.030733)

(0.023125)

(0.039546)

(0.029644)

LN Y

0.117183**

1.168891***

0.518018*

1.427659***

0.473954***

0.376770***

0.169181***

0.689775***

0.601621***

1.049838***

 

(0.055889)

(0.090563)

(0.269793)

(0.213328)

(0.083156)

(0.082948)

(0.064403)

(0.248777)

(0.150898)

(0.125336)

(LN Y)2

− 0.008562***

− 0.066846***

− 0.031274**

− 0.078420***

− 0.032316***

− 0.024336***

− 0.012537***

− 0.039248***

− 0.036843***

− 0.063656***

 

(0.003056)

(0.004929)

(0.014318)

(0.010656)

(0.004670)

(0.004394)

(0.003860)

(0.013111)

(0.008459)

(0.007124)

FIXCAP

0.001076***

0.007374***

0.005642***

0.005530***

0.001823***

0.004149***

0.001997**

0.007396***

0.006614***

0.004777***

 

(0.000399)

(0.000855)

(0.000966)

(0.000740)

(0.0003390

(0.000296)

(0.000780)

(0.001123)

(0.000963)

(0.000969)

TRADEBAL

0.000334**

0.001822***

0.000585

0.001049***

0.001158***

0.000516**

0.000955*

0.002814***

0.001093**

0.001821***

 

(0.000151)

(0.000490)

(0.000601)

(0.000297)

(0.000209)

(0.000239)

(0.000571)

(0.000717)

(0.000513)

(0.000651)

RIR

− 0.001737***

− 0.000529**

− 0.003029***

− 0.002873***

− 0.001207***

− 0.001734***

− 0.001101***

− 0.001625***

− 0.000666*

− 0.000431*

 

(0.000148)

(0.000251)

(0.000606)

(0.000433)

(0.000155)

(0.000219)

(0.000194)

(0.000460)

(0.000363)

(0.000246)

RER

0.000300***

0.000219*

0.000424***

0.000588***

0.000861***

0.000344***

0.000248**

0.000451***

0.000299***

0.000752***

 

(5.84E−05)

(0.000116)

(0.000100)

(0.000155)

(0.000127)

(5.35E−05)

(0.000110)

(8.51E−05)

(0.000110)

(0.000145)

INNOV*L_INCOME

 

0.055675**

9.18E−05**

0.025603**

9.29E−05**

9.22E−06**

4.51E−06*

0.008376***

0.020213**

0.001700***

  

(0.021625)

(3.71E−05)

(0.011596)

(4.08E−05)

(4.45E−06)

(2.32E−06)

(0.001503)

(0.008988)

(0.000620)

INNOV*M_INCOME

 

0.015431**

2.36E−05***

0.000147***

6.35E−07***

1.80E−07**

4.65E−08***

0.004364***

0.010358**

0.001006**

  

(0.006760)

(3.60E−06)

(4.02E−05)

(1.99E−07)

(8.43E−08)

(6.67E−09)

(0.000934)

(0.004720)

(0.000395)

INNOV*H_INCOME

 

0.012501***

7.48E−06***

3.64E−05***

1.12E−06***

4.01E−07***

3.11E−07*

0.002853***

0.011201*

0.000620**

  

(0.004466)

(1.65E−06)

(1.19E−05)

(4.19E−07)

(1.51E−07)

(1.65E−07)

(0.000885)

(0.006021)

(0.000291)

Obs

1016

482

393

313

701

836

710

366

549

583

Countries

78

53

44

41

72

68

64

53

58

65

Number instruments/number cross-section ratio

0.782

0.962

0.977

0.951

0.847

0.868

0.797

0.811

0.793

0.738

Prob J

0.160384

0.375879

0.332168

0.258393

0.384292

0.314723

0.301224

0.308133

0.174854

0.336980

AR(1)

− 0.420420

− 0.374270

− 0.402709

− 0.398945

− 0.354249

− 0.457684

− 0.439731

− 0.468025

− 0.389239

− 0.352414

P-value

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

AR(2)

− 0.052266

− 0.088957

− 0.103738

− 0.054130

− 0.069411

− 0.042501

− 0.050278

0.019809

− 0.043591

− 0.044065

P-value

0.1607

0.1135

0.1146

0.4427

0.1386

0.1915

0.2290

0.7711

0.4141

0.3586

  1. Information in brackets is the standard error associated with the coefficient
  2. Level of statistical significance: (***) denotes 1%, (**) denotes 5% and (*) denotes 10%
  3. S-GMM: based on Arellano and Bover (1995), two stages and no time dummy. AR (1) and AR (2) tests to verify the presence of first-order and second-order serial correlation in the waste in difference
  4. The number of instruments to the number of cross-section ratio needs to be higher than 1. Although the S-GMM estimates are adherent for samples with short periods and a high number of individuals, the diversity of instruments may generate the overlapping of instruments on the variables used, generating bias in the result (Roodman 2009)