IMPLAN Pro: Proving Analysis-By-Parts with a Comparison of Event Types

Background:

A traditional Industry Change analysis applies a change in demand or production for a sector in the input-output model. The direct effect is the value applied to the multipliers, the indirect effects are the inter-industry interactions, and the induced effects are the additional impacts caused by payroll as it is re-spent in the economy.

Analysis-by-parts (ABP), on the other hand, does not start with an impact on our target industry sector using an Industry Change. Instead we will break the analysis into two parts: Intermediate Expenditure impacts and Labor Income Impacts. For part 1, Intermediate Expenditures (IE), we will import an Industry Spending Pattern to specify the goods and services the target industry purchases in order to satisfy a demand or production level. The purchase of these goods and services from local sources actually represents the first round of indirect purchases by the target industry. For part 2, Labor Income (LI), we will use an Labor Income Change to analyze the impact of the payroll of our target industry necessary to meet the new demand or production level.

The purpose of this case study is to show that an analysis-by-parts (task B) can yield the same results as the traditional Industry Change analysis (task A). Then we will demonstrate a more realistic example of when an analysis-by-parts is useful (task C).

Task A. Traditional Industry Change analysis:

We wish to find the economic impact caused by a $2,000,000 expansion in Windows and Doors manufacturing. We will conduct the analysis first as a Industry Change analysis and will use the Washington County, MN 2007 demo data file found in every installation of IMPLAN Pro.

Steps:

1. Create the IMPLAN Pro model and build through multipliers.
2. Set the default year in user preferences to 2007: File > User Preferences > Analysis (tab) > and select "2007", then hit the "OK" button. This is so that when we analyze and compare the results for both tasks A and B, we do so without running afoul of deflators.
3. Create the "traditional" impact:
   1. Setup Activities > New Activity > (Activity Type) Industry Change. Name it 'Traditional' and hit the "Save" button.
   2. In the events window:
      1. Event Options > Show > Show All (this may already be done in your software)
      2. Choose "Sector" 99
      3. Set "Industry Sales" to 2,000,000 (note event year should be 2007 as a result of the user preferences).
      4. I know this is a corporation, so I am going to zero out the Proprietor income field. This sets the event to "custom" and causes red exclamation points to appear. This is what we want.
         

         Figure A1

4. Run the analysis: Name the Scenario "Traditional" and choose activity "Traditional".
5. Be sure the dollar year for view option on the Scenario Results screen is set to 2007.
   

   Figure A2

Task B. Analysis-by-parts:

We wish to find the economic impact caused by a $2,000,000 expansion in Windows and Doors manufacturing using analysis-by-parts (ABP).

We will use the same Washington County, MN 2007 model created in Task A.

Steps:

1. In order to make this work, we need to have all activity types available, so go to: File > User Preferences > Analysis (tab) and make sure that the "Advanced" box is checked. Check it if not.
2. Create the Intermediate Expenditures part of the ABP impact:
   1. Setup Activities > Activity Options > Import > Industry Spending Patter > choose sector 99 and hit the "Import" button.
   2. Click on the newly imported activity ("99 Wood windows and doors and millwork manufacturing") and notice:
      1. The events are coefficients and sum to less than 1. The missing piece is the portion of the production function that goes to labor income and other value added.
      2. The local purchase percentage ranges from 0 to 100%. In the traditional analysis, the LPP is 100% - that is all production of Wood windows and doors is local. We are now specifying as the "direct" effect the first round of spending in order to produce the windows and doors. We cannot assume that all those purchases come from local sources, and we let the software set the LPP for us. However, the analyst may change the LPPs if they have local information.
   3. Set the activity level for the intermediate purchases to 2,000,000. Since the production function shows the goods and services inputs required for each dollar of output, we set the activity level to the new production total being modeled.
      1. Click on the activity "99 Wood windows and doors and millwork manufacturing".
      2. Click on "Edit Activity" and set the Activity Level field to 2,000,000 and save.
3. Create the Labor Income part of the ABP impact:
   1. Setup Activity > New Activity > (Activity Type) Labor Income Change and name it "ABP Labor Income" and hit "Save".
   2. In the event window, hit "New Event" and choose Sector "5001 Employee Compensation". Set the Labor Income Value to the direct employee compensation associated with 2,000,000 in production. The value can be seen in figure A1 above: $690,031.23.
4. Create and run the new ABP Scenario
   1. Name the scenario "ABP" and include both the "ABP Labor Income" and the "99 Wood windows and doors and millwork manufacturing" activities.
   2. Compare the results to the traditional Industry Change analysis:
      

      Figure B1

Comments:

First, notice that the indirect and induced effects for both the traditional analysis (figure A2) and the analysis-by-parts (figure B1) match.

Second, why are the direct effects zero? By definition, any impacts resulting from labor income are induced effects. When an Employee Compensation Labor Income type activity is created, the software automatically moves the resulting first round of spending by households (the "direct") to the induced effects in the results. The software also moves any "indirect" effects resulting from the first round of induced effects to the induced category.

Likewise, the first round of spending on goods and services is moved from the direct category to the indirect category in the results. For an Industry Spending Pattern, there will still be indirect labor income, which generates additional induced effects. These induced effects are added to the induced effect generated by the initial labor income.

In the ABP analysis, the logical direct effect is the same as the direct effects specified and applied to the multipliers in the traditional Industry Change analysis (as shown in figure A2).

Related Topics:

The Basics of Analysis-By-Parts (Online)

The Basics of Analysis-By-Parts (Pro)