# Effect of Workers’ Remittances on Private Savings Behavior in Pakistan: Methodology

The use of an appropriate technique was considered as the center of any research study, without making a right choice for empirical analysis, the impact of study seems futile exercise. Therefore adoption of an appropriate technique was very important.

To analyze the connection between worker remittances and savings, one would need to use an appropriate model for empirical analysis. To examine the relationship between the variables in long and short run, many studies adopted the Johansen-Juselius and Engle Granger cointegration technique for checking the co-integration between variables in the long run. By using Johansen-Juselius technique, for co integration analysis, all variables deemed integrated at same order.

Therefore, to solve this issue, this study utilize Autoregressive Distributed Lag model (ARDL) developed by Pesaran, et al..This was a relatively new econometric technique used to estimate the relationship between savings and worker remittances and among other variables which are used in the study. This technique tests the co integration relationship without considering the same order of integration of all variables, either they were integrated of 1, 1 or mutually integrated.

Before following to other steps, the modeling strategy will be as follows:

1) Firstly the order of integration and stationary of the variables has been checked by using the Dickey-Fuller (DF) and Augmented Dickey-Fuller (ADF) test.

2) If the variables were integrated of the different order, then one can apply the ARDL (Autoregressive distributed lag model) developed by Pesaran, et al. to determine the co-integration for analyzing the long run relationship between variables.

3) If the variables showed the co integration relationship in the long run, then next one can apply the error correction mechanism (ECM) for analyzing the short run dynamics of the variables.

For checking the stationary of time series data unit root test is very influential. The time series which Many tests were applied to establish the integration order of time series in econometrics literature but most commonly was unit root test. Dickey and Fuller developed the unit root test. Augmented Dickey Fuller test is more favored over the Dickey Fuller because it gave more consistent results. The basic idea of co-integration is to identify the equilibrium in long run or relationship among the variables in long run. For illustration, the series Yt and Xt will be co-integrated of order I when the variables Y t and X t will be stationary at first difference I (Johansen and Juselius, 1992). By using Johansen-Juselius and Engle Granger techniques, in order to test for the co integration, all variables were integrated at same order.move slowly upwards and downwards around its mean with out any specific pattern are called random walks and the series which go upward and downward with a specific trend are called random walks with drift. installment payday loans