INCREASING THE EFFICIENCY OF TIME SERIES ANALYSIS USING THE ERROR CORRECTION MODEL (ECM)
Abstract: The study investigated equilibrium relationship of time series in the long-term and the short-term dynamics between agricultural employment labor and the factors influencing it during the period (2000-2016) and the degree of their integration so that accurate estimations can be found with a high level of confidence to depend on in the prediction with high precision. By examining the unit root test for time series stability, the time series for employment agricultural labor and explanatory variables is unstable at their level and at the first difference, while it was stable at the second difference (except the population was not settled and excluded).
The causal test showed that both the real daily wage of the worker and the real agricultural income caused the dependent variable (the agricultural employment). Therefore, the other variables that did not achieve the existence of the causal relationship between them and the dependent variable were excluded. The results showed co-integration relations between the agricultural employment and each of the real daily wage of the worker and the real agricultural income. Therefore, the standard relationships in both the short and long terms between these time series could be estimated through the so-called error correction mode. The results showed that all the estimates of the error correction coefficient were negative in the two models under study and less than the unity indicating their ability to quickly adjust to the balance of each of the explanatory variables, namely the average real daily wage per worker in pound and the net real agricultural income in billion and the dependent variable which is the agricultural employment that is considered a major problem in terms of scarcity and the migration of agricultural workers to other types of work such as working in the field of construction or industry which affects with damage to the agricultural activity.
The study recommends further studies to confirm these results, as well as directing research to study and analyzing time series behavior, and to conduct further research in the field of time series stability, causal relations, co-integration regression in the short and long term and error correction model (ECM) to reach least error in time series to increase their ability to prediction.
Publication year 2018
Organization Name
Author(s) from ARC
Publication Type Journal
↑