College: Modeling Public College & University Impacts with an Industry Spending Pattern using ABP



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. The purpose of this article is to provide an example of analyzing a Public College impact using Analysis-by-Parts with an Industry Spending Pattern to capture the university's Intermediate Inputs.


There is a new Event Type in IMPLAN called Industry Impact Analysis (detailed) released on October 6, 2021. In this Event Type, you can enter values you may have for Wage & Salary Employment, Employee Compensation, and spending on Intermediate Inputs (and also edit the Spending Pattern) or Output. Then, you can enter a zero for Proprietor Employment, Proprietor Income, Taxes on Production & Imports, and Other Property Income to reflect that public and nonprofit organizations are not paying these. And that's it. IMPLAN will calculate the Direct Effect for you! Check out the article Analyzing a Public College using Industry Impact Analysis (Detailed) to learn the easy way.


To perform an Analysis-by-Parts, you will want to know Direct Labor Income and the Direct Intermediate Inputs or Output, as well as the Industry that your business is best represented by. If Direct Labor Income, Intermediate Inputs, and Output are unknown, Output can be calculated using Employment and the Output-per-worker for the Sector that best represents your business or industry (Region Details > Study Area Data > Industry Averages, Output-per-Worker column, Industry row).  Hold onto this number, we’ll need it at the end. 

A handy template is here for you to use as you follow along with this example.




When the Industry you are modeling doesn’t match IMPLAN’s Output equation for the Industry, Analysis-by-Parts is appropriate. Performing an Analysis-by-Parts using an Industry Spending Pattern is most commonly used when modeling a nonprofit or public organization, because these organizations do not typically allocate an equal portion of their Output to Tax on Production and Imports (TOPI) or Other Property Income (OPI) as for-profit, private businesses. This technique is also useful for modeling a more specific type of industry than available in IMPLAN according to the default 546 Industry scheme, but this requires editing of the Industry Spending Pattern and the information to do so. 


Intermediate Inputs should be applied to an Industry Spending Pattern. These Industry Spending Patterns are made up of all the Commodities expected to be purchased by the Industry for annual operations. 

To analyze an Industry's Intermediate Inputs using an Industry Spending Pattern you'll need to follow these general steps:

  • Create an Event for the Intermediate Inputs.
  • Populate the Event with a Title that makes sense to you and for the Type, select 2018 Industry Spending Pattern. 
  • Choose the Sector that best matches the spending of the Industry/Organization being modeled.
  • By default the Value should be the total dollars spent on Intermediate Inputs.
  • Using the Advanced menu, you have the option to indicate that the Value should be Total Output, as well as make edits to the Commodities included in the Industry Spending Pattern.

Let’s take for example, a new public university that will open in West Virginia.  Select the Industry that best represents your business in the Specification Field. In this case, it is Industry 481 - Junior colleges, colleges, universities, and professional schools.  However, this Industry represents private establishments. 

Now, we need to put together the Output equation.


The Intermediate Input Coefficient can be found as the Gross Absorption column total in 

Region Details 
     > Social Accounts 
> Balance Sheets
> Industry Balance Sheet
> Commodity Demand

For WV in 2018 this value is 38.714%.


The additional coefficients for the Output equation can be found in the Value Added Coefficient column in:

Region Details 
     > Social Accounts 
          > Balance Sheets
               > Industry Balance Sheet
                    > Value Added


We know Output is the sum of Intermediate Inputs + Employee Compensation + Proprietor Income + Other Property Income + Taxes on Production and Imports. Therefore, these five numbers should add together to give you 100%.

IMPLAN says the Output Equation for this Industry in WV in 2018 is: 

Output (100%) = 

     Intermediate Inputs (38.71%) + Employee Compensation (41.58%)
     + Proprietor Income (1.55%) + Other Property Income (12.68%)
     + Taxes on Production and Imports (5.47%)

The public university we are hoping to model does not have any proprietors and it is tax exempt. While the public university does not generate any profit, it still may have some OPI for consumption of capital, this could include things like a purchase of new computers for the library. Determining how your industry or business’s Output Equation differs from IMPLAN’s definition is up to you as the analyst. 

If you have your own information about spending on Intermediate Inputs and Labor Income, this is ideal! Your Intermediate Inputs value can be entered into the Event Value Field.

If you do not have the total spending amount on Intermediate Inputs, you’ll need to calculate this using Output and your Industry’s Output Equation. 

To generate the Output Equation for the public university in our example, we need to remove PI and TOPI from our Output equation. We will do this by zeroing those out and renormalizing the equation.

First, we’ll need to sum the portions we want to keep:

     Intermediate Inputs (38.71%) + Employee Compensation (41.58%)
     + Other Property Income (12.68%) 
     = 92.98%

Next, we’ll divide each portion we are keeping by the new total to get new portions summing to 100%:

     New Intermediate Input portion: 38.71% / 92.98% = 41.64%
     New Employee Compensation portion: 41.58% / 92.98% = 44.72%
     New Other Property Income portion: 12.68% / 92.5% = 13.64% 

The value for Intermediate Inputs can be calculated as Output * Intermediate Input coefficient and entered in the Event Value field. If Output for the university is $5M, then Intermediate Inputs for the public university equals $5,000,000 * 41.64% = $2,081,912.73.


If you’d like to make any edits to this Industry Spending Pattern because some information about the universities operating expenses is known, you can do so by opening the Advanced Menu. 

This will open the following module where you can add Commodities (Arrow A), delete Commodities (Arrow B), edit the portion of Intermediate Inputs going to a given Commodity (Arrow C), edit the Local Purchase Percentage LPP (Arrow D), and re-sum the Intermediate Input coefficients to 100% by normalizing or convert back to the original settings by resetting (Arrow E). 


Editing an Industry Spending Pattern can become cumbersome if you have a full detailed list of expenditures.


Next, a Labor Income Event must be created either for Employee Compensation (EC), Proprietor Income (PI), or both.

In our example, the university will only have wage and salary workers so the total payroll value at the university should be run through a Labor Income Event with the Specification of Employee Compensation. 

Remember, these should be fully loaded payroll values which include wage and salary, all benefits (e.g., health, retirement) and payroll taxes (both sides of social security, unemployment taxes, etc.).  

If you need to convert your wage and salary data to fully loaded payroll, use the file Convert IMPLAN 546 Employment to FTE and Income to EC.

Let’s say in our example, payroll is unknown. We can calculate Employee Compensation by multiplying the university’s Output by the university’s Employee Compensation coefficient by taking Output $5,000,000 * 44.72% = $2,235,983.09.




Now either use the button at the top to select all or highlight each Event and drag them into your Group.  Next, hit run.




When your analysis is complete, the results will show you the economic impact of all of the Events you entered which will include the Industry Spending Pattern Event and the Labor Income Event.

In our public university example, the schools annual operations has Indirect and Induced impacts of approximately 19 jobs, $809,056 in Labor Income, $1.5M in Value Added, and $2.9M in Output.

  • Indirect Effects are from the Industry Spending Pattern Events only and represent activity in the local industries affected by Direct business’s supply chain.
  • The Labor Income events only create Induced Effects.


Because the spending by the Direct business was modeled instead of the Direct business itself, there is no Direct Effect in your Results. Using the information we calculated from Region Details, we can find the Direct Effects.

Define your Direct Effect:

  • We know Direct Output = $5M
    Also, IE + VA = Output

$2,081,912.73 + $2,918,087.27 = $5M

  • Direct Value Added = Employee Compensation + OPI (there is no PI or TOPI)
    $2,235,983.09 + $682,104.18 = $2,918,087.27
  • Direct Labor Income = Employee Compensation
  • Direct Employment
    Output/Worker = $72,279.89
    Output = $5M
    $5M / $72,279.89 = 69.18

To estimate the Direct Effects, take what information you already know for Employment, Labor Income, and Output and enter that directly. The template is on Recalculating Direct Effects will help you to estimate the remaining pieces based on data from:

> Region Details
     > Study Area Data
          > Industry Averages


Unfortunately cuts in public higher education are common. If the budget cuts are in general spending outside of payroll, this can be modeled using an Industry 481 - Junior colleges, colleges, universities, and professional schools. In this case, a negative value can be analyzed.

If you are interested in examining cuts to staff, we recommend modeling this using one of the special Industries at the end of the list: Industry 539 - * Employment and payroll of state govt, education. This will only give you Direct and Induced results, as no spending on Intermediate Inputs is included in these special Industries. A loss of 500 jobs in public higher education in West Virginia in 2018, would be modeled as shown.


When this Event is run alone, we will see the Direct and Induced effects of the loss of 500 jobs as well as the associated Labor Income, Value Added, and Output.




Recalculating Direct Effects


Hospitals: Modeling Private Hospital Impacts with Analysis-by-Parts

Hospitals: Modeling Public & Nonprofit Hospital Impacts with Analysis-by-Parts


Written July 3, 2019

Updated October 12, 2023