Aggregating expenditures

Hi Donna, Ideally, it would be much better to create a more detailed list of expenditures that would allow you to allocate a dollar value or even a percentage of the total expenditures that would go to each element of the spending. While you can create a summary multiplier as you have done, this would be a very rough approximation and we would not recommend this method, especially since it cannot take into account the relative weight (based on the size of the original expenditure) that each multiplier would have in the final results. Additionally, looking over you spending categories it appears that you are averaging together both the spending by the employees and your sectors of operations into the aggregated multiplier. You would definitely want to separate out the operations impacts from the spending of the workers in the local economy. Normally, you would run the labor income as a Labor Income Change Activity. However, if the workers on the film production program are not local (i.e., do not live in SW Colorado), you will not want to model their income via a Labor Income Change since they will likely take the majority of their payroll home. So you are correct that the best way to examine the Direct Labor component would be to treat it as a tourism study. Please let us know if you need some additional assistance.

Thank you. I also have a copy of the Seattle Film Study, and wondered how they arrived at their output multiplier (attached). Based on what the Office Manager has told me I think that 99.9% of the Film Crews that come in are non-local (though they might hire some local labor on occasion). I think that most of the expenditures made by visiting Film Crews are in Lodging and Food Services, though I believe they are also doing some discretionary spending at local stores. What if I "low balled" the average to reflect the average of those industries (as shown on attached) at 1.487099 so that output is not overestimated. Or would it be more sensible to just provide a range i.e. 1.369 to 1.696? I hate to belabor this but if the office manager cannot get detailed information from the crews then we cannot break it down by activity. I also want to make it easier for him to do the calculations by building multipliers into the spreadsheets. I think we can handle the local job impacts with the existing multipliers.

Hi Donna, Our suggestion would be to use the expenditures in Table 3 of the Seattle study to proportion out your known expenditure amount. I.e., sum up their non-labor expenditures and then divide each of their individual expenditures by that sum. Multiply this proportion by your known non-labor expenditure total to get an estimate of your expenditures on that particular item. Model each of these expenditures as an Event in a single Industry Change Activity (with LPP set to SAM Model Value). Then, after running the analysis you can calculate your own Output Multiplier by dividing your non-labor expenses by the total output impact. Let us know if you have further questions.

Most local labor used on film production is measured in days, so if we use job days as the input can we apply the multipliers of 346,347 or 348 (see attached spreadsheet - local labor multipliers)to get labor and labor income totals in job days? I am also including a short summary of how the film office director might interpret these numbers for his board or the general public.Any suggestions are welcome.

Hi Donna, We added a couple of questions/comments to your document and have reattached it here. There is no inherent problem with applying the Employment multipliers to hours; however, it appear that you might be double-counting with your current approach. It looks as if you are applying employment multipliers to the 3 types of direct labor, which will give a total employment impact (i.e., direct + indirect + induced), but then are also running the spending pattern, which will give indirect and induced related to that spending. In this case there would be overlap, and thus double-counting between these methods. If you want to use the spending pattern, then you can account for the Labor payments by running a Labor Income Change to get the rest of the induced impact associated to your direct labor payments and run the spending pattern and the Labor Income Change in a single Scenario. Then you can just add your direct jobs to the final impact - and not use the employment multipliers at all. Hope this helps. [attachment=403]SWRegionMultipliers_comments.doc[/attachment]

When I proportioned out the [b]non- labor[/b] expenditures from the Seattle study, and ran the numbers, part of the results were expressed in employment and labor income (see impact summary on Film Crew Spending worksheet), so I assumed that it was not directly related to film production. Is this assumption incorrect? That is why I thought we should look at employment multipliers directly related to film production (346-348) and apply them to (local employment) job days. Do you still think this might be double counting? Yes - I did combine direct and induced effects as it it sometimes hard for the client to separate/explain the 2 categories during presentations. I will include a table note that they have been combined into "indirect effects". Also - could we say something like "Motion Picture and Video Industries are found in IMPLAN sector 346 in the Southwest Region. The 10 direct job hours require about 2 supporting job hours (i.e. business support services, dentists, teachers, clerks in the local supermarket, or waitresses at the local restaurant)." Or is this too awkward? The Director rarely gets any solid wage info from local hires.

Hi Donna, I understand how this can be a bit confusing. While the non-labor expenditures are made in sectors other than the film sector, they represent the first round of the indirect effect associated with the film sector - it is just that you are modeling them yourself as opposed to using the film industry directly. These sectors that supply items to the film crew pay their own employees and buy their own supplies, which generates additional rounds of indirect effects. These are all part of the "Film Crew" impacts - they just happen to be indirect impacts. If you then run an impact on the film industry (which you have done by multiplying your direct employment by the film industry's multipliers), you will get direct, indirect, and induced impacts. But you've already modeled the indirect impacts above, so you are double-counting this portion. You have a few options: 1. You could use the multipliers as you have, but then omit the spending pattern portion. This is simple but loses some specificity by not making use of your known budget. Also, this will only give you employment impacts - no tax impacts, etc. 2. Use analysis-by-parts, of which you already have 2 of the 3 parts. With ABP, you model the indirect and induced effects separately, and then add to them your direct effects. You have already modeled the indirect, so now you just need to run a Labor Income Change Activity to get the induced effects. However, you state that you rarely get wage info, which will make this step more difficult. You could either dig for the wage info or you could use the Labor Income per Employee ratios for sectors 346-348 with your known employment values to get an estimate of Labor Income. Please see our ABP tutorial here: https://www.google.com/url?q=http://implan.com/v4/index.php%3Foption%3Dcom_multicategories%26view%3Darticle%26id%3D730:case-study-analysis-by-parts%26Itemid%3D71&sa=U&ei=naGfUZI14fjIAaiKgKgF&ved=0CAcQFjAA&client=internal-uds-cse&usg=AFQjCNEhFsyTeWUVin5r0b-CVe4SAZN76w Thank you for responding to our other questions - both ways you have suggested for reporting the impacts look good.

Thanks so much!