MRIO with Counties and States

Comments

1 comment

  • Avatar
    jenny
    Hi David, To get the most accurate results, you will need a State Package Data for the first state and State Level Data for the second stat. This will provide you with the state total and all the counties for the first state and the state total for the second state. This would allow you to perform true MRIO - the impact of activity in one region on another region. The easiest way to approximate MRIO would be to sum up all the final demands (from social accounts) for the county by commodity, subtract household PCE, and use that as your impact vector. You have to leave out households or they will be double-counted by the Type SAM multiplier. Alternatively, you could explicitly exogenize households from the multipliers when building the county and state models (File>User Preferences>Multipliers). Again, you would sum up all the final demands (from social accounts) for the county by commodity (this time without subtracting HH expenditures) and use that as your impact vector. However, in either case you will be missing the portion of households that derive income outside of labor income - e.g., retirement payments, government assistance. To calculate that value: [ol] [li]For each household row (10001 - 10009) in the aggregated IXC SAM, sum the row leaving out the payments by columns EC, PI and each of the household groups. [/li] [li]Divide that sum by the row total. That is the percentage of the Household FD that you will need to add back into your other FDs.[/li] [/ol] When you omit a county, you are losing all the final demand generated by that county. Your analysis will still be conservative, since some of the imported final demands will come from the rest of the state.
    0
    Comment actions Permalink

Please sign in to leave a comment.