## Steps to conduct Causality Analysis

Analysis of interplay of economic factors in any economic system is always of the interest of the economic analysts across the world, and across several domains of economics. One of the major analytical tools in analysis of this particular kind of economic aspect is the causality analysis. This is one of the complex analytical tools being used in economic analysis, and it involves a series of sequential steps. First, in order to carry out a causality analysis, it is needed that the variables under consideration are stationary in nature, and most of the variables used for economic analysis are flow variables in nature. By conducting unit root test, stationary of the variables are ensured. Second, if the variables under consideration are not associated on a long run basis, then the causality analysis can never prove out to be significant. Therefore, in order to investigate about the long run association among the variables, co-integration test is conducted on them and based on the Trace and maximum Eigenvalue statistics number of co-integrating vectors among the variables is calculated. Presence of at least one co-integrating vector ensures the presence of long run association among variables. Third, once the long run association among variables is finalized, as a final step it is needed to find out the directions of possible causal associations. This step is carried out by applying vector error correction methodology, and by formation of vector autoregressive equations involving the variables under consideration. If the error correction term in any equation is significant, then long run causal association exists from the independent variables to dependent variables. If the individual coefficients of independent variables are significant, then short run causal associations exist from those variables to the dependent variable. This is the summary of causality analysis tool in economic and econometric analysis. For more information about conducting a causality analysis, kindly browse through the pages of...