I have been conducting analyses at both state and regional levels. I am using the same template for importing regional activities (different values for each activity in each state). When I sum the DIRECT results for the individual states, I get totals that are approximately 10% greater than the value that I get when I run the regional model. I have done this for several Census regions, and am seeing the same results. Why is that? I would expect the state-level DIRECT effects to sum to the same or close to the regional results.
Please sign in to leave a comment.