Skip to main content

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

Table 8 Nonlinearities of undervaluation and service sector performance.

From: Real exchange rate undervaluation, regional integration and services sector performance: evidence from the East African Community

 

(1)

PMG-ARDL

(2)

PMG-ARDL

(3)

CS-ARDL

(4)

CS-ARDL

Speed of adjustment

 ECT

− 0.208***

(0.026)

− 0.318***

(0.092)

− 0.624***

(0.115)

− 0.942**

(0.358)

Longrun

 ln Capital

  

0.458***

(0.113)

  

0.274**

(0.071)

 ln Labour

  

− 0.262***

(0.081)

  

− 0.030**

(0.953)

 Growth

  

− 0.007*

(0.004)

  

− 0.001*

(0.047)

 Inflation

  

− 0.006***

(0.002)

  

− 0.003**

(0.007)

 ln Undervaluation

8.275**

(3.350)

6.916**

(3.041)

10.853***

(1.993)

10.270**

(5.303)

 ln Undervaluation2

− 4.947**

(2.177)

− 4.296**

(2.089)

− 6.644***

(1.325)

− 5.951**

(3.974)

Short run

 ln Capital

  

− 0.079***

(0.012)

  

− 0.242**

(0.114)

ln Labour

  

0.083

(0.206)

  

0.043

(0.207)

 Growth

  

− 0.000

(0.002)

  

− 0.008

(0.011)

 Inflation

  

0.001***

(0.000)

  

0.000***

(0.001)

 ln Undervaluation

− 4.040

(3.958)

− 3.126

(2.632)

− 0.129

(0.334)

0.716

(0.424)

 ln Undervaluation2

2.253

(2.400)

1.642

(1.570)

3.519

(3.990)

− 0.501

(8.673)

Constant

0.077***

(0.018)

0.152***

(0.054)

1.437

(3.506)

4.215

(4.895)

Observations

104

 

104

 

104

 

104

 

Diagnostic tests

 Hausman test (χ2, p-value)

1.23

[0.5404]

2.87

[0.476]

− 

 

− 

 

 Wooldridge Autocorrelation test

6.504 [0.269]

18.119 [0.2381]

      

 Normality test

0.94 [0.625]

3.75 [0.1531]

      

 Heteroscedasticity test

121.27 [0.241]

25.09 [0.1110]

      
  1. The Hausman test was carried out to choose between Mean Group ARDL and PMG-ARDL, and the latter was preferred. “ln Undervaluation2” is the square of the log of undervaluation; All variables are in log form except Growth and inflation; Standard errors in parentheses
  2. ***p < 0.01
  3. **p < 0.05
  4. *p < 0.1; ARDL (1 0 0 0 0 0)