can the national data base provide this info

What you propose will certainly tell you which sectors, in producing for final demand, have the most US backward linkages - if that is your definition of a healthy US economy. You will need to avoid double counting; for example, production of assembled automobiles will require a tremendous amount of automobile parts. To enter the entire production of both Assembled Autos and Auto parts will double count the parts production and all associated backward links. You have two choices to avoid such double counting: 1) Enter only the production delivered to final demand. 2) "Break the links" by manufacturing to the other manufacturing by setting the RSC (amount of local supply going to local demand) to zero for the manufacturing sectors. In the US (or any model) it is certainly possible to change the amount of imports coming from outside the region by setting the RSC for the commodity of interest. However, since the amount of US goods sold to US demand is limited by the amount of US goods, you will need to do the analysis by increasing US imports, with the assumption being that the amount of economy supporting local production is the same whether or not it is a gain or a loss. However, changing the import rate only changes the multiplier and is not in itself an impact. It would probably be simpler to apply a production change to the industry producing the commodity as an impact based on how much the value of production would have to increase ($'s not %'s) if we were to import less of it. The "Bill of Final Demand" at the US level is ultimately controlled to the BEA NIPA accounts for the year of the data. The distribution of those demands to industries comes from many sources. Details can be found in the knowledge base here: http://implan.com/V4/index.php?option=com_multicategories&view=categories&cid=241:datainformation&Itemid=10. To answer your questions: 1. The absorption and by-products tables are based on the latest BEA benchmark (currently 2002). The absorption matrix is updated using a “RAS” to the current Industry Output, Institutional Sales, Final Demands and Value Added at the US level and controlled to the BEA NIPA accounts. 2. The default Type SAM multiplier endogenizes the labor income (proprietor income and employee compensation) by respending that income through the Household institution. The type SAM multiplier can be changed using the IMPLAN software to endogenize any of the institutions. 3. The structural matrices which are based on the 2002 BEA benchmark as noted in 1, are projected to the current year. The rest of the study area data is for the current year; or if lagged a year, projected to the current year. All of the data is controlled to the US NIPA accounts for the current year of data. 4. We have sold data sets for other countries, but as they have gotten older they are no longer listed on the order form. We are setting up a project to create data sets for many countries based on the data from the OECD and Eurostat social accounts models. I missed a question - the software will aggregate the 440 sectors in the US data sets to any aggregation you desire based on a template you create through the software.

Doug/jeff leeson, thanks for your thorough reply. The definition of a healthy US economy, for my dissertation, is defined as the real GDP growth rate...so yes, the number/depth of backward/[forward?] linkages will determine impact on GDP. Will IMPLAN output give me some kind of new total GDP after I make my changes? I'm not quite sure of your meaning about trading imported mfg'd good for US made...I may come back with a f/up question. Thanks and have a good holiday

Doug/jeff, here's one f/up question pertaining to your import substitution response. "However, changing the import rate only changes the multiplier and is not in itself an impact. It would probably be simpler to apply a production change to the industry producing the commodity as an impact based on how much the value of production would have to increase ($'s not %'s) if we were to import less of it." Your very last suggestion sounds good and easy but won't that then increase total GDP which I'm trying not to do? Don't I have to somehow "withdraw" import demand? jeff

Impacts are created by applying a change in production/employment for specified sectors. Increasing/decreasing imports would be simulated by setting new production levels (change) for the industries producing those commodities. It is possible to affect the multipliers by editing underlying social accounts matrices but results won't be generated until you apply changes to those multipliers. Changes in value added (GDP) and all the components of value added are part of the impact results generated by the software.

Decreasing imports requires increasing local production (and GDP) to meet existing demand and vice versa.

Doug/jeff, ok, to your very last answer.....lets say I just assume, for simplicity that X% of manufacturing sector final demand, is currently imported. Further lets assume its proportionately distributed among the Final Demand "Bill of Demand". So I could increase the production levels for the mfg sectors...I get that.....but will software then know that this means X% less must be imported? In other words is IMPLANS $ of imports a backed into number, vs. a "built up number" from some data base?

Running an impact (e.g., simulating an increase in mfg output) does not change the underlying SAM and thus does not affect the import rate of manufactured goods. However, that does not matter. Import substitution is equivalent to an increase in local mfg output and if you model this increase in mfg output as an Industry Change activity, you will get all the associated backward linkages of this increased output. This is an increase in final demand and is a classic I-O setup.

Jenny/Jeff,happy holidays... 1. your use of SAM in your first sentence, is this what you mean to be the "Final Bill of Demand" 2.I'm sorry but I do not understand the rest ....maybe I'm using the wrong terminology...if I substitute local mfg for imported mfg'd goods[without increasing final total domestic demand/consumption], this increases local linkages, employment,GDP....right? My question is still this: If all I do is increase national mfg by X%[to replace some imports], but leave final consumption/demand constant, will IMPLAN itself, know to decrease what had been the imported requirements to meet that final demand...or do I need to tell IMPLAN something or do you mean its a moot point?

Input-Output models (the underlying technology of Implan software) create a complete set of accounts for the economy - ie, the SAM. The SAM shows the flow of goods to and from each of the sectors/factors/institutions of the economy. Included in the purchases by industries and institutions are the estimates of how much is bought locally versus importing. The multipliers are derived from that picture with the assumption being that any new demand will require spending in the same pattern as demonstrated in the SAM. An increase in production applied to the multipliers (the process of impact analysis) will increase the requirements of the local economy to satisfy that demand - with the associated implications to GDP. We can look at the Commodity Summary Table (under Explore>Social Accounts) and see that for commodity X: "Net Commodity Supply" (net of foreign export) = 100,000,000 "Total Gross Commodity Demand" = 200,000,000 "Average RPC" = .25 (That is 25% of the 200 million - ie, 50 million is met by local supply, therefore 75% or 150 million is imported). We then ask; what if we were to replace the 150 million (might as well shoot for the moon) of imports with local production? How would the economy respond - ie, what are the total requirements of economic activity needed to satisfy this new 150 million of local production. To calculate the impact we apply 150 million to the multipliers.

Doug/jeff leeson, A couple more questions 1. mechanically how are the separate input, and output tables combined into the total requirements table 2. how is the final "bill of demand" derived? Simply by decomposing our GDP by industry? 3. I'm amazed that the total multiplier for the manufactuirng sector of about 2.4 is only marginally different than that of agricultures of 2.35[2002 BEA data]. Intuitively I really struggle with that. Is it because the final bill of demand has a disproportionate share of food vs "manufactured" products...or ????? Thanks

1. In the manual (link below) the second book (pages 131-174) show how we regionalize the US absorption and by-products matrix, remove imports, create the industry by industry transactions matix (although we now use the gravity model trade flows rather than econometrics to derive the RPCs) and the various methods to create the multipliers (total requirements). The type SAM multiplier with the Households endogenized is the default methodology for creating the multipliers in IMPLAN. http://implan.com/V4/index.php?option=com_docman&task=doc_download&gid=105&Itemid=7 2. So rather than increasing US output by a certain amount to replace a portion of US import, you wish to change the entire structure of the US economy, to assume, for example 100% of manufacturing is local? To create multipliers using this assumption would require editing the US model study area to increase the size of the industries' output to provide the capacity to meet gross demand with local (US) production. This would require several steps: a) Set foreign imports to 0, this would have to be done by editing the US model through MSAccess as the software does not allow you to do this. This will result in The "domestic" demand for commodities to increase correspondingly and, with the current balanced US model, will result in "domestic" import as the US capacity to satisfy the increased local (US) demand does not exist. b) edit the US model (Customize>Study Area Data) by increasing the value added, output and employment proportionately to satisfy the "domestic import". c) Repeat step b a number of times as increased output will also increase domestic demands. You may be able to project the final output needed after several iterations. The resulting model will ideally minimize the domestic imports and domestic exports (ie, be balanced) and the multipliers will reflect the targeted foreign import. The multipliers will be changed but an impact will still have to be run; eg, a vector of manufacturing final demand applied to the US model before modification and after modification - with the difference reflecting the changed structure. 3. I can only speculate, that the output per worker for manufacturing is extremely large, meaning that the induced effect (meaning the labor income/output ratio is relatively small) is small compare to agriculture. I would imagine that the indirect effect is comparatively large for manufacturing which would be why Manufacturing still has a larger output multiplier.

Another option to accomplish your purpose is through re-balancing the 2002 BEA benchmark: 1)Remove the desired negative import from the import column 2)Calculate the new TCO (total commodity output) - either the sum of the Use matrix row plus final demands row or current TCO minus the negative import removed. 3)Run the TCO through the Market Share matrix to calculate the new required total industry output (TIO) 4)Increase the value added elements proportionately based on the change in TIO. 5)Run a RAS on the Use matrix so that it balances with the new TCOs, TIOs and value addeds. From here you create the multipliers (Excel does have the tools to do this): 1) New Use (with Households if creating type ii) times market share to create the industry x industry transactions. 2) Create the A matrix by dividing the Use (with employee compensation if creating type II) by TIO 3) do the matrix algebra 1/(I-A) to create the Leontief inverse. I have not seen employment for the 2002 benchmark - you could obtain it from the 2002 Census. BLS covered employment and wages would be a proxy but it does not have complete coverage nor include proprietors and differs significantly from Census for a number of sectors.

Doug and others/jeff leeson, I've finally begun working on the models to show how important our mfg sector is to our economy. To begin, I tried adding 1 million of sector 276 auto mfg to see what the output looks like; I see total VA of 971 and output of 2668 resulting from the added 1 million of spending. Is the 2668 equivalent to GDP$? If so, and to calculate impact on GDP growth rate, what would I use for my "base" level of GDP before any modelling? 2. I then began to try some import substitution by going into the trade flows tab. I note the total RPC of .607173 for this industry. I tried to adjust that to a higher figure to show more US mfg/less imports of cars but I kept getting an error msg of "demand exceeding supply". I did not get this msg when I simply upped auto mfg by $1 million so I assume it has something to do with the what I'm trying to do with the RPC? In an earlier msg to me you suggested adding something to atuo production manually at the same time as I decrease imports but I'm not clear how to do that 3. on a separate note, is there a "flow chart" diagram of how your different pages are linked together. For example is the RPC something I can adjust or is it a backed into number from whatever US production and import $ happen to be? The flow diagram would be useful to know WHEN I have to manually adjust a few things vs when adjusting one thing automatically updates other things 4. if I wanted to model the ENTIRE economy growing X% at the same ratios of purchases[PCE %'s], how might I do that? Thanks

Hi Jeff, [ol] [li]Value Added = GDP. Output is the total value of production and includes the GDP number. You can pull the base level of GDP from the Model Overview screen.[/li][br] [li]At the national level (i.e., in a U.S. model), RPC is constrained by the Supply/Demand Pooling ratio - that is, you cannot purchase more local commodity than is locally produced. To increase the RPC above this, you would need to increase the supply of the commodity by going to Customize > Study Area Data. However, modeling the impact of import substitution does not require such editing. If we are importing less of a commodity while demand for that commodity stays the same, then U.S. production of that commodity would have to increase to cover the difference. This can be modeled via a simple Industry Change Activity. [/li][br] [li]The Flow Chart is setup in how you read through the menus (i.e., if you are going to modify Study Area Data, Industry Production, Commodity Production, and Trade Flows, the modifications have to be done in that order). Editing the Study Area Data (e.g., increasing output of a particular commodity) will require you to rebuild the multipliers, which in your case will automatically adjust the RPCs since you are using a U.S. model, which uses the Supply/Demand Pooling multiplier method. If you edit the output of a sector and rebuild the multipliers, the RPCs will automatically adjust so long as you are using the Supply/Demand or Econometric multiplier methods. So you will not need to make this adjustment manually. However, as stated in 2, you do not need to make any adjustments to Output – you can simply run an Industry Change Activity. [/li][br] [li]Could you explain exactly what you mean by “the same ratios of purchases[PCE %'s]”? Do mean that Household spending increases by a certain percentage?[/li] [/ol]

Thanks...what I mean about #4 is the same "final bill of demand" meaning "purchases" remain in the same relative ratio[products/industries]...I keep referring to this is the PCE

Reply #2 from jeff: On your answer #2, am I correct in saying when I adjust the us produciton UP per your advice, that the RPC will automatically go down as a result? thx, jeff

I tried "you do not need to make any adjustments to Output – you can simply run an Industry Change Activity. " with no luck; I keep seeing screens with coefficients but no screen where I "add" $ of us production in that industry; what menu option and what screen should I be in? Thanks

Hi Jeff, When you run and Industry Change Activity, or any other type of Activity it will not affect the underlying Model. This is why you don't have to worry about keeping Final Demand a constant. You can just Model the value of your % increase or decrease and it will show you how local production with respond as a result of that additional production in the U.S. or will decrease as a result of increasing imports. So in regards to your last post, you are correct, this won't affect the Study Area. The regional purchasing coefficient increases as more of US supply is used to meet US demand. An RPC of 1.00 means all local demand is met by local supply. Again, while the Industry Change will show you the associated PCE spending (Induced impact) resulting from the production or decrease in production, since this doesn't affect the overall final demand in the model you would just be seeing the purchases associated to the increase in local production that you are modeling. This increase is substitutive of what was imports in the past.

thx...I've since sent you another msg of how to do this "industry change" activity; I'm lost