I am doing a multi-regional analysis of each Hawaiian county on a base model I named "west coast" with Washington, Oregon, and California. What is the difference between combining these states initially, and linking them to a model as one region, versus linking each individual state (Washington, Oregon, and California)? I compared the results in both instances (for the model with individual states, I summed up the total effects for each state) and they were slightly different, although not too far off. Do these differences represent the economic activity between the states not captured in the grouped model?
Please sign in to leave a comment.