Abstract
This study applies Granger causality tests within a multivariate error correction framework to examine the relationship between judicial caseload, real income and urbanization for Australia using annual data from 1904 to 2001. Decomposition of variance and impulse response functions are also considered. The Granger causality results as well as the decomposition of variance and impulse response functions suggest that urbanization is the most exogenous of the three variables in both the long run and short run while judicial caseload and real income are relatively exogenous in the short run.
Acknowledgements
We thank Olga Ryan and Xiaolei Qian for research assistance with this project and an anonymous referee for several helpful suggestions.
Notes
1 Note that this result is not sensitive to the choice of deterministic trend. Cases 1, 3–5 in Johansen (Citation1995) were also tested and these cases suggested a single long-run relationship. Results are available on request.
2 Urbrain (Citation1992) and Harris (Citation1995) explain that the exogeneity test in a cointegrating framework must be based on the significance of the error correction term in the short run model. As a result, exogeneity is tested for by examining the statistical significance of the error correction term. Johansen (Citation1992) notes that this procedure satisfies both conditions of weak exogeneity as explained in Engle et al . (Citation1983). It is found that when income and caseloads are independent variables, the error correction terms are statistically significant, implying that one can reject the null hypothesis of weak exogeneity. However, the null hypothesis of weak exogeneity cannot be rejected when urbanization is the independent variable.