ESTIMASI PARAMETER MODEL REGRESI NON STASIONER DENGAN VARIABEL DEPENDEN LAG : STUDI KASUS PADA PERKEMBANGAN EKSPOR INDONESIA KE JEPANG TAHUN 1980 - 2000

Abstract

The clasiccal regression model was devised to handle relationship between stationary variables. But, many economic variables that frequently faced by econometricians when dealing with time series data, are nonstationary variables. This clearly places severe restrictions on their analysis by standard regression method.

In this paper, we study regression models with a lagged dependent variable when both the dependent and independent variables are nonstationary, and the regression model is not cointegrated. In particular, we discuss the limiting properties of least squares estimates of the parameters in such regression models. We show that the estimate of the lagged dependent variable tends to unity and the estimates of the independent variables tend to zero. The results might also allow us to investigate the growth of export from Indonesian to Japan.