multiplier analysis
I am comparing multipliers in a geography. Is there a way to calculate the aggregated multiplier for say Ag., Mining, Mfg., etc? Second, if I insert a new supplier industry in to a geography, is there a way to calculate the new multiplier for the industry sector which uses those inputs or a way to estimate the affect the new industry has on the geographies multipliers. My concerns is that if I add a key local supplier industry, that I will be understating the multiplier for related industries.
Finally the multiplier (2010 Model) for insurance agencies and brokers is quite high in the geography I am studying. Is it the employment and salaries of staff that create such a high multiplier 1.91 ~ 2.07? This seems more like a manufacturing multiplier. Thoughts?
-
You can build an aggregated model to get the aggregated multiplier. If you add a sector, when you rebuild the model, most of the other multipliers will change due to the interaction of the sectors. If you do add an industry, edit the trade data so there is some local demand for the new commodity produced by that industry. Which multiplier are you talking about? Output? What region is it? -
OK so new sector to the geography changes the multiplier, but only if one changes the trade flows? If I add another industry, say a new insurance agency into a geography where there is already insurance agencies, will that change the industry multiplier- NO or very small amount of multiplier change for industries using brokers as a commodity? If I add a machine shop to a geography, where none is current, how do I show that some of the local production will be local? Will that then change the multipliers of ALL industries which use the machine shop commodity- YES? We are trying to figure out which industries have the highest multipliers, but that is dependent on local supply right? But I speculate this is not so easy, since introducing a new industry into the mix could significantly affect ALL existing industries multipliers (not just most important) if they purchase machine shops parts locally. We are trying to figure out which local industries would be most affected by a new outside industry coming into the geography. Is this a matter of looking at the USE table and selecting the commodities not present, then add to the geography, making some of that local production available to the already present industries. For machine shop how does one edit the trade data to impact the multipliers? Finally, we need to identify industries which have the highest multipliers, without listing all 400 + thus the aggregation question. When I aggregate then just do a calculation of the multiplier based on total output / by direct. Thank you. -
Hi Scott, The tradeflows (and thus RPCs) from the National Trade Flows Model method are fixed and do not change to reflect changes in supply or demand, so you would need to manually change the RPCs yourself (Customize > Tradeflows). You can do this either for a. Individual users of the commodity: Click on the Industry/Institution tab, select the commodity from the drop-down menu, then edit the individual purchasing sectors' RPCs as you see fit or b. All users of the commodity: To edit a commodity's RPC for all industries and institutions, you have to edit the RSC in such a way that results in the desired average RPC. The definitions and equations below can be combined for this purpose: Local Use Ratio (RSC) = the proportion of Net Local Supply that is used to meet local demands = (Local Use of Local Supply)/(Local Net Commodity Supply) = (Local Use of Local Supply)/(Total Supply – Foreign Exports) RPC = the proportion of Local Demand that is met by local suppliers = (Local Use of Local Supply)/(Local Domestic Commodity Demand) RSC = LULS / LNCS and RPC = LULS / LDCD so newRSC = (desiredRPC * LDCD) / LNCS Note: if you use the Econometric RPC method, the RPCs will adjust automatically to the new supply/demand once you reconstruct the multipliers (i.e., no need to make the manual changes above). If you aggregate the model, you do not need to calculate the aggregated multipliers yourself - the multipliers report (Explore > Multipliers) will reflect the new aggregation so long as you re-constructed your multipliers after you aggregated (Options > Construct > Multipliers). Regarding the seemingly high multipliers, which type of multipliers are you looking at - e.g., Output multipliers, Employment multipliers,...? -
Jenny thank you. Doing a normal impact will not impact the multipliers, the impact just uses the multipliers which are available, unless I create a new sector, not already in the study area. In order to note this change I either need to use an econometric model or make the trade flow changes manually. In order to change a multiplier I need to change the trade flow data, making more (less) of the trade "local". I tried this and it does change the multiplier. Why is it necessary to use the econometric RPC model versus the normal national trade flow model? The formula is fine. I assume that one would use this if there is a specific piece of analysis one needs to accomplish since it is very detailed. Aggregation thank you, that works. Finally, we are trying to identify the industries which would have the largest change in multipliers if there were an economic shock to the study area. Would these typically be the sectors which utilize the greatest percentage of their inputs from local sources? The method that we are discussing requires the analyst to review each sector. Is there a method for narrowing the list of sectors which would increase (decrease) their multipliers the most based on a change in the study area? The large multipliers were for a single county, insurance agencies (brokers) sector 358. The output multipliers were 1.91. Again this was part of looking for sectors which would have the largest potential jump in a multiplier if there was a change to the local economy. That multiplier seemed high, since they do not purchase anything, but likely do have large salaries and all their work 99% is local (RPC 99. 1). So in this case there would not be much of a change, only down, if I were to decrease the percent local RPC. -
A sector's multiplier is an indicator of its inter-connectedness to the local economy and depends on how much of its inputs are purchased from within the study area and how much it spends on inputs and labor income vs. going to profits. So a sector’s multiplier will change if: a. It changes its input structure b. Some of its input suppliers leave the study area or move into the area c. It changes the amount of Labor Income it pays per dollar of Output. Commuting rates would also be expected to affect multipliers, but IMPLAN has just a single industry-average commuting rate for a given study area. If you have specific commuting rate information for your sectors of interest, you can customize the impacts for these sectors to reflect your known commuting rates and then calculate their multipliers based on total impact / direct impact. One thing to keep in mind is that even if a sector has a high multiplier (i.e., greater local impact per $1 of output), if that sector's total output is very small, it may not be that significant to the local economy. So it is always good to include context when reporting multipliers and may be useful to report expected output growth/contraction along with expected multiplier changes. Regarding the high multipliers for sector 358, I would agree with your thoughts about high labor income. Also, while this sector does not buy a lot of manufactured inputs (most of which would have been imported from outside the study area data anyway), it does buy a lot of services, more of which can probably be sourced locally. To see what the sector buys (gross absorption) and how much of it is sourced locally (RPC), go to Explore > Social Accounts > Balance Sheets tab > Commodity Demand tab and select sector 358 from the drop-down menu. If you click on the “Gross Absorption” column heading, it will sort according to this field. Remember, the RPC for sector 358 itself does not give an indication of how much it purchases locally; rather, it indicates how much insurance is purchased locally, which is a forward linkage that is not reflected in sector 358’s multiplier (but is nonetheless a very useful metric when determining a sector’s importance to the local economy). -
Jenny, agree on all points. Do you have a source the for formula which you provided, book or white paper? Last paragraph- The trade flow RPC is then an average of the RPC of the gross absorption of the commodities used by the industry? Attached speadsheet (single county) shows what I am running into. Trade flows RPC is 99 percent but the commodities in the balance sheet do not indicate that. I am missing a piece here, what do I not understand? Thank you. -
Hi Scott, We don't have a paper for that equation. It's just that in V2 it was possible to directly edit a commodity's RPC for all purchasers, whereas in V3 you have to take the round-about way of editing RSC to affect the change on RPC. So the formuala tells you what you need to set RSC to in order to get your desired RPC. A commodity's RPC says nothing about where the providers of that commodity purchase their inputs. In your example, commodity 3358 (Insurance agencies, brokerages, and related) has an RPC of 99.96 in your region. This says that of all local demand for insurance agencies and brokerages, 99.96% is met by local insurance agencies and brokerages. It says nothing about where those insurance agencies and brokerages (i.e., sector 358) purchase their inputs. So you could say something like, "99.96% of our region's demand for insurance is met by local agencies and brokerages, and those agencies and brokerages purchase roughly 30% of their inputs locally."
Please sign in to leave a comment.
Comments
9 comments