Issues with model results

IMPLAN, After running my model scenarios, I found that the results seem to be...errant. Here are the details, I'll explain the problem after: Data files: three West Virginia counties, 2010 Models/Activities run: Current impact, future impact Events run for each Activity are based on correlation with categories of tourism expenditures as categorized by the US Forest Service. The correlated activities are: 328, 411, 410, 326, 410 (again), 413, 330, 338, 324, 412. The values I input for each activity represent tourist expenditures, by category, within the study area in 2012. Therefore, I changed the Model Year to 2012. I used the default "Industry income/values" as the input category. I then ran each scenario (current, future) separately. The results for each came out seemingly wrong. The reason I say this is because the Output results ended up being less than the expenditure/income values I entered as event input. This would suggest a total multiplier of less than 1.0. However, when I examined the multiplier results, the "total" multipliers for each associated industry were all greater than 1.0. Secondly, I ran the "current impact" model a second time to see if the same issue came up, and it did, but not only that, the results came out differently. I am certain that before running each model I added up my input values to make sure that the total reflected the total in my source spreadsheet, and it did. And the second run resulted in an even lower Output than the first run. I'm uploading the multiplier spreadsheet from the second "current impact" run. I've added a tab with the data I input and the results from running the model the first time. Even though the attachment shows the multipliers from the second run (while the results are from the first run), the multipliers from the first run were also all greater than 1.0. In the first tab (multipliers) I've highlighted the sectors/codes/events that I used. Thoughts? Thank you for your assistance.
Was this post helpful?
0 out of 0 found this helpful


1 comment

  • Hi Rory, Thanks for all the great information it really assists us in being able to work out what's going on. Based on your description and information, it looks like two things are at play here. The first is Margins. As is common with retail studies there are several categories/Sectors that you are using that are retail Sectors, and it looks from your results like you correctly applied Margins to these purchases. When you apply Margins, the software takes the entered retail value and reduces it to represent on the portion of sales that is kept by the retailer to operate their business, and then leaks the remaining value. A typical retail Margin is between 20 and 30%, but you can see exactly the percentage and the values that are actually being applied to the Multiplier from your initial Industry Sales value in two ways. If you are just interested in the value, you can click on the Preview button in the upper right-hand side of the Setup Activities screen. This will show the values as they are applied to the Multiplier adjusted for both inflation (based on your selected Event Year) and the Margined value. The second method uses this pathway [url=]Event Options> Edit Event Properties>Margins> Edit[/url]. This shows you both the Margin percentage and the Margined value. This occurs because the Purchaser price (what you and I pay) is much larger than the retailer's markup, the only part of the sale that retailer actually keeps and uses for operations. The remaining dollars entered into Industry Sales are leaked from the Model, because we don't know the basket of goods purchased. Without knowing what was purchased, we can't attribute the purchases of the actual items to any particular producing Sector, determine whether or not it was produced locally (which is typically unlikely) or what portion of its wholesale or transport costs may be local, so we have to assume that all production or the purchased product, transport, and wholesale are imports. The issue you are seeing in regards to different results in different years is likely associated to the deflators. If you are summing results, you will want to choose a single Dollar Year for View to sum them in, but if you are viewing and reporting them relative to the year of the expenditures you will want to be sure the Dollar Year for View field matches your Event Year for the Event. You can check your sum of entered values at the bottom of the Event screen in the "Sum of Event Values". This sum does not include any adjustments for Margins of deflators so it will help you ensure that you have entered the desired value into both Activities.

Please sign in to leave a comment.