Industry Output Event vs Industry Impact Analysis (Detailed)

Hello – I hope this inquiry finds you doing well! I have a couple of questions that pertain to a methodological approach I’m ironing out for an economic contributions analysis.

I’m estimating the impacts of a business operating in the IMPLAN sector of 445 – Insurance agencies, brokerages, and related activities. The geography is Douglas County, Oregon.

The intermediate inputs are $2.434 million for 2021 (I am using 2021 IMPLAN data too). Of the total intermediate inputs, about 49% ($1.212 million) are locally purchased; the remainder (51%, or about $1.222 million) is purchased outside of Douglas County (i.e., elsewhere in Oregon or in another state). Employee compensation is estimated to be about $2.176 million for 32 FTE (all within Douglas County).

I am testing a couple different modeling methodologies so I can better understand the nuances of each and ultimately settle on the proper methodology to use. My questions are as follows:

  1. “Simpler scenario”. I ran an Industry Output event with Total Output = $3.388 million (sum of employee compensation = $2.176 million and local intermediate inputs = $1.212 million). The resulting output SAM multiplier was about 2.6 under this scenario, which seems far too high for a single county. Given that the intermediate inputs I modeled were filtered for local expenditures (i.e., Douglas County) only, is it appropriate to add back in the non-local intermediate input expenditures to the direct output (after running the model) to generate the correct SAM multiplier? (If I add back in the $1.222 million of intermediate inputs not captured in this model, the SAM multiplier is about 2.2, instead of 2.6, which is still quite high).
  2. “More complex scenario”. I have access to the NAICS sectors of each local expenditure made by the entity in question. Under this scenario, I tried an “Analysis by Parts” (ABP) approach where I generated a unique spending pattern for the entity. The inputs in my Industry Impact Analysis (Detailed) scenario are as follows: 32 Wage & Salary Employment, 0 Proprietor Employment, $2.176 million Employee Compensation, $0 Proprietor Income, $0 TOPI, $0 OPI, and $1.212 million in Intermediate Inputs. The total output is about $3.388 million (same as the “Simpler Scenario”). With these inputs, the resulting output multiplier is about 2.0. However, as far as I understand, when generating results under the ABP framework, I need to add back in the non-local expenditures I did not model (to properly account for the entity’s full business expenditures, local and non-local). If I add back in the non-local expenditures ($1.222 million) into my output results column, the multiplier is about 1.7, which seems like a much more reasonable multiplier relative to the “Simpler Scenario”. Does this ABP approach seem correct in the way described? If not, could you please provide me some insights on how I should model the entity properly under an ABP framework?

Given that I have access to the entity’s line-item business expenditures, I think the ABP framework makes the most sense to use. However, I wanted to inquire about my approach under each model run to glean any additional insights on how I could properly recalibrate my inputs.

Please let me know if I can clarify my language/thoughts in any way, and I will happily do so. Thank you so much!

Was this post helpful?
0 out of 0 found this helpful



  • Official comment

    Hello Ryan! 

    Couple of clarifying remarks here: Analysis-by-Parts is a technique by which you can analyze the impact of an Industry's production/spending in separate components using multiple Events instead of using a single Industry Event. This is not the same as an Industry Impact Analysis (Detailed) or IIA Event, where you can enter Employment (Wage & Salary and/or Proprietor), Labor Income (Employee Compensation and/or Proprietor Income), Intermediate Inputs (and make edits to the Industry Spending Pattern), Taxes on Production & Imports, Other Property Income, and Output from the Impacts screen in a single Event. In fact, the IIA Event Type was developed to alleviate the necessity of utilizing the ABP approach in IMPLAN Cloud. 

    With that being said, I do agree that the IIA Event is going to be best for you here. I would say that you should enter the full value of Intermediate Inputs, and account for the locality using the Local Purchase Percentage for each Commodity in the Spending Pattern. Even though these purchases will not create additional local impacts, they are a part of the Direct Output value and should be included accordingly. Additionally, it is important to note the IMPLAN Employment is not exactly the same as an FTE, and there is a necessary conversion process to convert your FTE/wages to an IMPLAN Employment/ EC value

    Lastly, I am curious why you have zeroed out Proprietor Income/Employment, OPI, and TOPI? We would only recommend this in the case that you know definitively that these are in fact 0, otherwise, you should allow IMPLAN to estimate those values based on the underlying data. 



    Michael Nealy

  • Hey Michael!

    Thank you so much for your prompt response to my inquiry! I appreciate your note re: ABP, as I had misunderstood the essence of the IIA event type (i.e., I wrongfully assumed IIA was a stand-in equivalent for ABP). I will be mindful of this going forward.

    Also, thank you for the note regarding FTE. I was aware of the conversion workbook, but I thought I could input my Wage & Salary employment as FTE, run the model, and then convert the indirect/induced FYE results into FTE using the workbook. However, given your note, I see the error I've made; I will fix that accordingly. As a small addendum, the Wage & Salary dollars I'm using in this analysis include salaries, payroll taxes, and fringe benefits. Are those three components of my W&S dollars accurate to model as W&S? Or, does it make more sense to allocate the payroll taxes to TOPI?

    Re: the 0s, I do know there is no proprietor income nor proprietor employment. However, the 0 values I assigned to TOPI and OPI are erroneous given your insight. I will re-run my model and allow IMPLAN to estimate those values (assuming I don't place payroll taxes in the TOPI box).

    If I can clarify any of my above remarks, please let me know! Thank you so much again for your time and insights.


    Warm regards,


  • Hey Michael,

    Upon further reading of IMPLAN's income data (, it looks like payroll taxes are captured in Employee Compensation. Given your note about converting FTE to IMPLAN Employment, do you think it makes sense to convert my FTE estimate, but keep my Employee Compensation estimate the same (i.e., do not use the Wage & Salary-to-EC conversion factor)? I ask because my EC data already includes salaries, fringe benefits, and payroll taxes, so I assume I am representing EC correctly in the model.

    Thank you again!


    Kind regards,


  • Hello Ryan!

    Just to be explicitly clear here, Employee Compensation in IMPLAN is the total payroll cost of wage and salary employees to the employer. This includes wages and salaries, all benefits (e.g., health, retirement) and payroll taxes (both sides of social security, unemployment insurance taxes, etc.). It is also referred to as fully-loaded payroll. If you believe that the compensation value you have includes all of these items, then I would agree that you are all good to input your default data as the Employee Compensation value.



    Michael Nealy


Please sign in to leave a comment.