Abstract:

Overviewed in this paper are important and quite recent approaches used in estimating long-run relationships. A cointegration technique is largely used in macroeconomic modelling. This technique simply provides an effective formal framework used for estimating, testing and modelling long-run relationships from time series data. Particularly, described in this paper is the estimation procedures of the techniques which have received more attention in the recent applied literature. This paper explores the use of a Vector autoregressive model, augmented by an error correction term to model the gross domestic product - agricultural production nexus. Time series data spanning 1990 Q1 to 2014 Q4 sourced from the South African Reserve Bank and the Department of Agriculture are used. Exhibited by the model is an insignificant model showing a long run relationship between the variables over the past 13 years. The ECM confirms that for the system to go back to equilibrium, about 80% of adjustment may be effected per year. No causal relationship between the variables has been reported. Due to the findings urgent attention must be paid to agricultural sector, lest we find ourselves with no food in the country. The government is urged to farmers with farming resources to ensure the viability of this sector. Emerging farmers must also be equipped with proper farming skills. Not only will this boost the economy of the country, this may also help eradicate rather high unemployment and poverty levels in the country. High food prices may be reduced and better living standards improved.