Table 1.
Descriptive statistics of variables.
Table 2.
Correlation matrix of variables.
Table 3.
Unit root at first difference.
Table 4.
Pedroni residual co-integration tests.
Table 5.
Model 1 (NSDP) unrestricted co-integration rank test (trace value & maximum eigen value).
Table 6.
Model 1 (Per Capita NSDP) unrestricted cointegration rank test (trace value & maximum eigenvalue).
Table 7.
Panel two stepsGMM using NSDP as a dependent variable.
Table 8.
Panel two-step GMM using NSDP/Capita as dependent variable.
Fig 1.
Residual, fitted, and actual values of NSDP.
It is graphical residual representations along with actual and fitted values NSDP &NSDP/Capita using the GMM model. As we can see the residual or error terms of the model presented through blue color are within the limit of -1 to +1. While actual and fitted values of NSDP are moving alongside. There is not much deviation between the actual and fitted values of NSDP. Table 8: Panel two-step GMM using NSDP/Capita as dependent variable.
Fig 2.
Residual, fitted, and actual values of NSDP/capita.
It is demonstrates graphically the residual, fitted, and actual values of per capita NSDP. As we can see the residual or error terms of the model presented through blue color are within the limit of -0.4 to +0.4. While actual and fitted values of NSDP per capita are moving alongside. There is not much deviation between the actual and fitted values of NSDP. Considering the residual, actual, and fitted values above two GMM models can be considered as good fit models and these further can be used to forecast NSDP and per capita NSDP.
Table 9.
Fully modified ordinary least square (FMOLS) estimation model-1 (NSDP) and Model-2 (Per Capita NSDP).
Table 10.
Growth rates of states for selected indicators.
Table 11.
Rank of states based on growth rates for selected indicators.