As discussed previously, the author does estimation and sees that saving ratio, long run growth and the population are drifting together. From this view, it can be suggested these three things can be co-integrated. The author suggests establishing co-integration, first thing first need to check whether each series is integrated and contains a unit root. Dickey-Fuller has been improved with a lag and the constant in the equation of this test are shown below:
S/Y change of S/Y g
ADF t-stat -4.0*** -0.47 -0.55
Change of g M/E change of E/M
ADF t-stat -3.3** 2.75 -2.75*
***:1 %, **:5%, *:10%
Other than that, MacKinnon critical values for rejection of hypothesis of a unit root.
The tests show that the three series contain a unit root. The author tests whether the series are co-integrated over the sample period and if so, what the co-integrating relationship is. They use two methods which is the Engle and Granger (1987) two-step method. There are two stages in this model. First stage is consists of regressing the variables supposed to be co-integrated, while the second stage consists of testing for a unit root in the residuals. The second method they use the basic model.
First and foremost, they see growth of S/Y is not very smooth but instead is interrupted by several “bumps” that represent short-term, transitory deviations from the long-term trend. Based on LCH or the permanent income hypothesis, the propensity to save during such transients should be exceptionally high.
Then the author tests the impact on the saving ratio. According to them, the impact is depends on the definition of income and saving. From their view, they measured inflation by the quantity pA*/Y. They conclude that according to their model, the parameter –k, measuring the impact of inflation on S/Y could be expected to be...