An Estimation of the Long Run Relationship between Agriculture Output, Export and Investment in Sindh-Pakistan (Some Evidences From Johnson Model 1988-95)

Primary goals of any agricultural investment policy must be increasing productivity to protect food security and promote economic development of the region by export. The main goal of this study is to evaluate total investment and its co integration with agriculture output and agriculture export in Sindh-Pakistan. This study concludes that investment is helpful in the short run and long run. More investment of public and private sector  will boost  export of agric commodities in era of free trade will also do more to raise productivity and thereby promote agricultural and economic development in emerging economies like Sindh –Pakistan  in the long-run.

The close inspection of graphs shows the existence of strong time trend in the data and coefficients of the sample autocorrelation die down at high lags and indicate that these series are not stationary in levels and confirms the presence of time trend in the series. The ADF test (1979) is also concluded that all variables are I (1) in their levels and I (0) in the first difference at 5 percent level of significance. The result of granger causality test (1969) revealed that none of the above combinations has bivariate causality. The result shows the cointegration relation exists in lagout, lagexp, linv, lpinv for Pakistan though it is unidirectional causality not bivariate. However for Sindh the cointegration relation exists in lsagout, lsinv and lspinv. The multivariate cointegration test of Johansen’s (1988) concludes that both trace and eigen vale test indicates one cointegration relationship between lagout, lagexp, lginv and lpinv.  However for Sindh only trace test indicates one cointegration relationship between lsagout, lsagexp, lsginv and lpinv.