Impact analysis

This sounds like a classic case for a contribution analysis. This forum discussion describes the topic of contribution anlysis and contains a link to detailed instructions for conducting one: http://implan.com/v4/index.php?option=com_kunena&func=view&catid=84&id=10897&Itemid=35#10902 Please let us know if you have additional questions.

Thank you for the suggestion. I'd like to try running the contribution analysis, but I'm not sure what to do with updating the industry data that IMPLAN shows. Due to primary data I have, I already know the employment generated by tenants being able to use the airport and I know the employment at the airport, but I'm still wanting to see the induced and indirect impacts. I think I need to customize the Study Area Data for Industry 332, Transport by Air, to show my information, but I'm not sure how to handle the employment generated through the tenants' use of the airport. I assume that altering the Study area data to show only employment attributed to the use of the airport will cause me to lose the impact of those jobs that aren't associated with the airport. Any suggestions on how to handle this situation would be appreciated, thank you.

Regarding specifying your known employment figures, this can be done when setting up your Events - simply override the value that IMPLAN puts in for you. You will see a warning message - this is fine, it is just alerting you to the fact that it is now a customized value. So long as your air transportation figures don't include the tenant figures, you can run each of the tenants as a separate impact in addition to the airport operations. If you are doing this as a contribution analysis, you have already set RSC for air transportation to 0, so there will be no double-counting of impacts to air transportation. If not, then to avoid double-counting the impact to the local air transportation sector, first go to Customize > Trade Flows > Industry/Institution RPC tab, then select commodity 3332 from the drop-down menu. Set the Regional Purchase Coefficient to 0 for any of your tenant sectors. You will need to reconstruct your multipliers after the change.

Thank you for your reply. I have a question regarding developing my Contribution Analysis: If I set RSC to 0 under the 'Trade Model' tab, do I also need to customize the RPC to 0 for the tenant sectors under the 'Industry/Institution RPC' tab? I'm just curious if both are necessary, or if I should one or the other. I'm assuming I should only modify the RSC because the RPC's automatically change when I modify Commodity Production. Thank you for your time.

That is correct - only the RSC needs to be changed. If none of the local supply is used locally (RSC = 0) then none of the local demand can be met by local supply (RPC = 0).

Thank you for your reply. I am using industry 338 for my contribution analysis and I'm wondering if it will work fine to set 3332, "Air transportation services", as my primary product within 338. I would give it the coefficient of 1 instead of 3338, which is currently the primary product. My reasoning for this is I expect it would exclude the contributions of Rail, Water, Truck, and sightseeing products. Will that work correctly? Thank you for your time.

If you are trying to see the economic impact of air transportation, I would use sector 332. Even if you change sector 338's commodity production to be solely 3332, this will not change sector 338's production function (i.e., the inputs and labor purchased by the sector), so it will have very little effect on your analysis results. Remember that I-O only looks at backward linkages. So the only differences would be the feedback effects to sector 338 stemming from sector 338's input purchases (e.g., suppose through the additional rounds of indirect and induced, someone has a demand for Scenic and sightseeing transportation and support services, they would no longer be able to purchase it from sector 338).

Ok, I understand. There are a few reasons I wanted to use 338. Firstly, the NAICS codes I'm wanting my contribution analysis for are more accurately represented by 338. These include things such as aircraft maintenance, fueling services, hangar rental, cargo handling, or airport operators. All of these are listed under industry code 338. At least this is the impression I'm left with when I do a sector search. Secondly, when I go to modify the coefficients I am being told that 332 doesn't exist in the Region I'm studying. But, 3332 exists within 338. Thank you again for your time.

Ah yes, you mentioned in your first post that the sector did not exist in your region - I had lost sight of that. You can make it exist by using the data you have and national averages to fill in any blanks: 1. Go to Customize > Study Area Data and scroll down to sector 332. 2. Enter anything you know into their respective boxes, then click "Update" and "Save" 3. Reconstruct the multipliers: Options > Construct > Multipliers 4. Proceed with the analysis as before.

Thank you for the reply. I am planning to have two activities in my contribution analysis. One will be for the industry I'm running the analysis on, with the Industry Sales equal to the Total Output. Then my other activity includes an event for each sector represented. For these events I'm entering the known employment. I want to confirm that I'm doing this part correct. When I've read through the Contribution Analysis process it said that setting the RSC to 0 is done so that no one purchases from my industry beyond what I specify in the event setup. Does that mean that I create a second activity, then set up events specifying what is purchased from my Contribution Analysis Industry? Thank you for your time.

Am not sure what you are trying to accomplish with the second activity. Is this activity of the sector you are trying to model - or secondary industries associated with a primary activity (eg, freight services that use an airport)? When you specify the output of an industry as an impact, the multipliers of that industry already contain the spending by that industry of local goods and services. To specify this separately as another activity double counts. Can you tell us specifically what you are trying to model in this study?

You may find this discussion helpful: http://implan.com/v4/index.php?option=com_kunena&func=view&catid=84&id=11178&Itemid=35#11233

I want to know the contribution of the an airport, so I'm trying to run a contribution analysis on industry 338. But I also have other information that I want taken into account in the analysis. This other information comes from tenants at that airport who credit a given number of jobs to being able to use the airport. I was directed to this page, http://implan.com/v4/index.php?option=com_content&view=article&id=660%3A660&catid=253%3AKB33&Itemid=1, which details how to run a contribution analysis. In it I am told to set RSC to 0 for industry 338 so that no other industry can purchase from 338 unless specified in the event setup. In the contribution analysis should I just have one activity that would include events for 338's Industry Sales as well as one event for each sector represented by the tenants? Thank you for your time.

Thank you for clarifying. This does indeed look like a multi-industry contribution analysis. You would have an activity for the airport and each of its tenants. For each of the industry's represented (i.e., the airport and each of the tenants), you need to set the RSC to 0 and reconstruct the multipliers before any impacts are run. However, I would again suggest using sector 332 (after you make it exist as discussed previously). Sector 338 includes support activities for other types of transportation (water, road, other) and the activity is local in nature, usually involving a same-day return to the point of departure. If you go to Explore > Social Accounts > Balance Sheets tab > Commodity Demand tab and select sector 332 from the drop-down menu, you will see that sector 332 actually makes a purchase of commodity 3338 - so it is purchasing these air transportation support services and you will still see an impact on sector 338 by running your analysis on sector 332.

Thank you for the reply. I don't think I need a multi-industry contribution analysis, and here is why: the study I'm doing is to show the contribution of an airport, not an airport and its tenants. The businesses and people that use this airport are tenants that credit a percentage of their employment to that specific airport. That percentage correlates to a certain number of jobs and those jobs are a contribution of the airport. In effect, I have an industry for which I'd like to see a contribution analysis, but I also have data for that industry's direct contribution. In other words, I know the jobs credited to the airport for the study area, but I want IMPLAN to give me the Indirect and Induced effects associated with the direct effect that I already know. My understanding for how to do this is to: 1. Set RSC to 0 for Industry 332, which will not allow any purchases from 332 beyond what is specified in the event setup, then 2. Specify the purchases in 332 by creating an event for each industry represented by a tenant at that airport. Will this work? I hope explained better this time, thank you for you assistance and patience.

I think I see now - so you will run an event for each of the tenants, but each of these events will be in sector 332 (after making sector 332 exist and setting its RSC to 0)? That sounds good.

I was thinking I needed to put the tenant events in their respective sectors. For example, I would show an event for a given number of jobs to represent a glass manufacturing company that uses the airport and I would put it as sector 158. Is that not correct? Also, would it go into the same activity?

That was my previous understanding - and this would be a multi-industry contribution analysis as suggested in post #12151. I think the confusion is arising from the term "multi-industry" contribution analysis. While you are not interested in the contribution of the entire manufacturing industry, for example, but just the portion that stems from the airport's existence, the procedure necessary for modeling this is the same as the instructions for a so-called "multi-industry contribution analysis". The key is that you also need to set the RSC for the tenant sectors to 0. These can be run as separate activities or together in one activity - it is user preference and will yield identical results. You can then run all the activities as a single scenario and avoid having to sum the results in Excel (unless of course you are interested in the separate contributions of each activity, in which case, you would run them each as a separate scenario).

I want to quickly confirm my process for this contribution analysis, simply because the my numbers seem so large on the scenario results. I've attached an Excel file with my results. The Direct Labor accurately represents what I entered in the event setup. 1. Altering study area data for 332 only if my primary research has different information than IMPLAN currently shows. 2. I'm making certain the GRP doesn't change throughout the process. 3. I'm modifying commodity production. I don't understand why, but 3332 is the only product and therefore the primary product listed under 332 when I go to modify commodity production. Hence, I don't have to do anything in this step. 4. I would reconstruct the multipliers if I'd changed commodity production. 5. Modify trade flows - I'm setting RSC to 0 for 3332. 6. Reconstruct multipliers. 7. Confirm Changes - recheck GRP, look at Foreign and Domestic Exports, check intermediate output. 8. Create event for 332 equal to Industry Sales. 9. Create event for each tenant's sector, filling in their reported job numbers. If you see anything missing from this process please let me know. Thank you for your time.

If you are truly wanting to run this as a contribution analysis (i.e., you want the total impact on the airport sector and its tenants to equal their current activity), as opposed to an impact analysis (which would show additional feedback effects to the airport sector and its tenants), then you need to follow the same steps 3-6 for the tenant sectors as well. The way you have it set up now, you have combined a contribution analysis of the airport with an impact analysis on the tenants. The way to check your results is if the total effect on the air transportation sector matches its current level of activity. Ditto for the tenant sectors if you do end up modeling those as a contribution analysis too. Not all sectors produce more than one commodity, which is why you didn't end up having to make any changes in step 3 - that's okay.

Thank you for the reply. When checking my results, should I be looking at the total effect on output? Also, is the current level of activity something that IMPLAN tells me, or does that information come from somewhere else? Thank you

In the "Detail Results" tab, the Total Effects for Output, Employment, and each component of Value-Added for the air transportation sector (and the tenant sectors should you decide to run them as contribution analyses too) should equal their current values found in Explore > Study Area Data. The Total Effects across all sectors provides the complete contribution of these industries to the model area.

Is Total Output the same as the Output column found in Explore > Study Area Data? My Direct Employment comes out low when I use the figure from the Output column. Thank you for your time.

You want to compare the Total impacts (not the Direct) to the Study Area Data.

"That is correct - only the RSC needs to be changed. If none of the local supply is used locally (RSC = 0) then none of the local demand can be met by local supply (RPC = 0). " Hi Jenny, I'm curious about your reply to Tim's question about RSC vs RPC = 0. Tim, are you running a trade flows model or econometric RPC? I have run a contribution analysis (econometric RPC model) with only RSC=0 and the same model again with both RSC and RPC = 0, and the second model's (both =0) results were smaller. Can you explain when you mean it is appropriate to only set the RSC to 0? I understand you mean theoretically, if the region doesn't supply then the region can't purchase a good, but the results were different, which makes me think I don't completely understand. Thanks!

When using the tradeflow RPC method, there is no way to change RPC directly, which is why you need to change RSC. When using the econometric RPC method, it is the other way around. If you are using the econometric RPC method and you can see the "Trade Model" tab, you will notice that the RSCs are all already zero - this is because this table is not used in the econometric models - so setting them to zero here has no effect.