Page 17 - CITN 2017 Journal

the model is accepted. The next test is to construct the Error Correction Model (ECM) to
determine the speed of adjustment of the short run and the long run analysis. This is
represented in Table 3.

4.2.3: Error Correction Model
Table 3: Error Correction Model (ECM)
Dependent Variable: GEH(1)
Method: Least Squares

Variable Coefficient Std. Error t-Statistic Prob.
C -3431.943 6591.390 -0.520671 0.6083
MREM 0.103762 0.034448 3.012117 0.0069
FDI(1) -0.014437 0.002875 -5.022516 0.0001
EGS(1) 0.017927 0.003088 5.806457 0.0000
FX(2) 253.1541 100.0286 2.530819 0.0199
ECM(-1) -0.3629 247.4706 -0.742026 0.0467
R-squared 0.981913 Mean dependent var 96365.81
Adjusted R-squared 0.977391 S.D. dependent var 131340.9
S.E. of regression 19748.91 Akaike info criterion 22.81876
Sum squared resid 7.80E+09 Schwarz criterion 23.10909
Log likelihood -290.6439 F-statistic 217.1484
Durbin-Watson stat 3.047458 Prob(F-statistic) 0.000000

Source: Authors' Computation, 2016

At this juncture, it is pertinent to know that the explanatory variables used in the study are
all statistically significant. It is established at this point that migrant remittances from
abroad have positive effect on the economic development of Nigeria. The error correction
term coefficient has the expected negative sign (-0.3629). This is indicative that 36 per cent
of the discrepancy between the short-run and the long-run equilibrium value GEH is
corrected or eliminated annually. That is, a speed of adjustment of around -0.3629 is
required for equilibrium to be attained in the long-run. Should there be disequilibrium in
the economy, it will require an adjustment of about 36 per cent in the long-run to take place
either by market mechanism, government intervention or both.

4.2.4 The Ordinary Least Square
Table 4: Regression Result
Dependent Variable: GEH
Method: Least Squares

Variable Coefficient Std. Error t-Statistic Prob.
C -2586.654 4323.328 -0.598302 0.5544
MREM 0.050721 0.011239 4.513021 0.0001
FDI -0.011056 0.001724 -6.414015 0.0000
EGS 0.021153 0.001801 11.74572 0.0000
FX 234.3862 83.93633 2.792428 0.0093

R-squared 0.987930 Mean dependent var 92529.99
Adjusted R-squared 0.986206 S.D. dependent var 146693.9
S.E. of regression 17229.08 Akaike info criterion 22.48531
Sum squared resid 8.31E+09 Schwarz criterion 22.71206
Log likelihood -366.0076 F-statistic 572.9492
Durbin-Watson stat 2.956125 Prob(F-statistic) 0.000000

Source: Authors' Computation, 2016

10