Table 5

Globalization, inclusive tertiary school education (TSE) and structural transformation

Dependent variable: manufacturing value added (MVA)
Globalization (kofgi)Economic globalization (kofecgi)Social globalization (kofsogi)Political globalization (kofpogi)
 MVA (−1)0.994***0.951***1.045***0.897***
(0.067)(0.059)(0.088)(0.088)
 TSE0.016−0.797−0.375−0.674
(1.029)(0.978)(1.303)(0.893)
 Global0.140***   
(0.049)   
 Global × TSE0.106***   
(0.037)   
 Econ global 0.071  
 (0.061)  
 Econ global × TSE −0.092  
 (0.061)  
 Social global  0.152** 
  (0.063) 
 Social global × TSE  0.136** 
  (0.051) 
 Polit global   0.060*
   (0.035)
 Polit global × TSE   0.073**
   (0.030)
 Population growth1.199***0.807**0.874***0.790**
(0.283)(0.322)(0.293)(0.297)
 Time effectsYesYesYesYes
 Constant1.1193.696***1.0033.882**
(1.751)(1.160)(1.216)(1.636)
Net effects0.064na0.0550.008
Thresholds1.320na1.1170.821
 AR (1)(0.034)(0.033)(0.027)(0.033)
 AR (2)(0.654)(0.927)(0.791)(0.735)
 Sargan OIR(0.877)(0.365)(0.775)(0.446)
 Hansen OIR(0.880)(0.537)(0.933)(0.767)
 DHT for instruments    
 (1) Instruments in levels    
 H excluding group(0.709)(0.287)(0.484)(0.222)
 Dif(null, H = exogenous)(0.812)(0.649)(0.969)(0.948)
 (2) IV(years, eq(diff))    
 H excluding group    
 Dif(null, H = exogenous)    
 Fisher136051.34***31282.02***16488.57***45880.90***
 Instruments34343434
 Countries36363636
 Observations263263263263

Note(s): “***, ** and *, respectively, denote the 1%, 5% and 10% levels of significance. DHT shows the Difference in Hansen Test used to assess the exogeneity of instruments subsets. OIR: Over-identifying Restrictions Test. Dif: Difference. Italic values have two principal significances: on the one hand, the significance of the Fisher statistics and estimated coefficients; on the other hand, the non-rejection of the null hypotheses of (1) autocorrelation absence in the AR(1) & AR(2) tests and (2) the instruments that are valid based on the Sargan and Hansen OIR tests. na: not applicable given that at least one estimated coefficient that is indispensable for the computation of net effects does not reflect significance. The mean of tertiary school enrollment is 0.712. The range of tertiary school enrollment is 0.064–1.494. For the estimated coefficients, values in parentheses reflect standard errors, while for the information criteria (i.e. AR, Sargan, Hansen, DHT and IV tests), p-values are disclosed

or Create an Account

Close Modal
Close Modal