INTRODUCTION
Examining changes in the overall economy can be a helpful tool in economic development. Using IMPLAN, you can see what not only a change to certain Industries can be, but you can also examine questions like what an overall economic decline or increase would look like.
DETAILS
Let’s examine a 10% decline across the entire U.S. economy. These examples use Industry Aggregation with the 2-digit NAICS scheme so that there are only 22 sectors, however, the same can be done for all 546 IMPLAN Industries. There are three different options when setting up the analysis of the Output loss across each sector.
Industry Output Events
This will capture the 10% decline across all sectors as the Direct Impact and will capture the Indirect and Induced impacts to sectors supported by the direct loss. Analyzing a 10% loss in Total Output in each sector produces a Direct Output of -$3.7 trillion (10% of Total Output in the US in 2018), but the Total Output results from the analysis is $9.2 trillion. That is an overall 25% decline in US Output! Keep in mind Industry Output includes production going to Final Demand and Intermediate Output. Intermediate Output includes production sold to other Industries to be used as an input in their own production. If the estimated decline being analyzed includes expected supply chain effects and/or effects from losses in wages, running an Industry Output Event could overestimate the impact. However, if a decline in a subset of the economy is being analyzed, such as a 10% decline in household industry demand, then there are additional losses to be expected that can be analyzed via Industry Output Events based on 10% of the existing total household demand to each Industry (Region Details > Industry Accounts > Reports > Household Industry Demand).
First, we will navigate to Region Details in our Region. Scrolling down to the Industries table and hovering in the upper right corner, download the table.
Now in Excel, calculate what a 10% decline in Output is for each sector. The total loss would be a staggering $3.7 trillion!
Next, we jump into our Impacts screen and enter one Event for each of the sectors with the value for a 10% decline in each (column F). Note: there is no Output for sector 21, so there is no need to enter an Event for it.
Then drag the 21 Events into your Group and hit Run.
Navigating to the Results, the Direct Output is $3.7 trillion, which yields an Indirect Output of $2.6 trillion, and an Induced Output of $3.0 trillion, for a total potential loss of $9.2 trillion.
On the Output tab, we can see that the sector hit hardest by this scenario is 31-33 Manufacturing with a loss of $1.6 trillion and the sector affected the least is 61 - Educational Services with a loss of $64 billion.
Industry Contribution Analysis Events, run in a single Group
This will capture just the 10% decline across all sectors. Results will not include any Indirect or Induced Effects because this Event Type restricts purchases to the specified Industry and all other Industries specified in an Industry Contribution Analysis Event with the same Group.
In this example, we will model the same 10% decline across all sectors using the Industry Contribution Analysis Event Type. There are two ways to set up Industry Contribution events, using the dollar value or percentage value. Instead of using the dollar values from the first example, we will set up the events as a -10% value across all sectors.
Then drag the 21 Events into your Group and hit Run.
Navigating to the Results, we see the same Direct Output of $3.7 trillion from the first example. However, these results do not include any Indirect or Induced Effects because this Event Type restricted purchases to the specified Industry and all other Industries within the same Group.
Industry Contribution Analysis Events, run in separate Groups
This will capture the 10% decline across all sectors and will capture the Indirect and Induced contributions to sectors supported by the direct loss. Results for each Event will not include Indirect or Induced Effects to the Industry specified in the Event because this Event Type restricts purchases to the specified Industry and all other Industries specified in an Industry Contribution Analysis Event with the same Group.
Finally, we will model the same 10% decline across all sectors using the Industry Contribution Event Type, but run in separate groups. We set up each sector as a -10% Industry Contribution Event, same as the second example. However, in this case we need to create a new group for each sector, by using the menu icon to duplicate the US Total group. We should end up with 21 groups, the same as the number of events. Then drag each event into its own group. Notice that the first group only has the Agriculture sector and the second group only has the Mining Sector, and so forth. When all groups are populated with an event, hit Run.
Navigating to the Results, the Direct Output is $3.7 trillion, which yields an Indirect Output of $1.7 trillion, and an Induced Output of $2.3 trillion, for a total potential loss of $7.7 trillion. These results are slightly less than the first example because we restricted purchases back to the specified Industry within the Group. Since each sector was separated into its own group, the purchases were not restricted back to the other remaining sectors as in example 2.
Find more information about Industry Contribution Analysis here to learn more and determine if an Industry Contribution Analysis Event is right for your study.
RESULTS COMPARISON
When comparing the results, the first thing we notice is that our direct effect values are identical across all three analyses. The difference occurs in calculating the indirect and induced effects.
To examine a 10% decrease in output of the US economy, we first used Industry Outputs Events, which resulted in the largest indirect and induced effects, as no restrictions were made on buybacks to the selected industries. In the second option, we restricted all buybacks to all industries in the group resulting in zero induced and indirect effects. In the third option, we restricted buybacks only to the industry itself, which resulted in a more conservative estimate of indirect and induced effects than the Industry Output Events. This option resulted in a total difference of 6.9 million workers, $451.5 billion in labor income, $795.0 billion in value added, and $1,569.3 billion in output.
| Event Type | Effect | Employment | Labor Income ($B) | Value Added ($B) | Output ($B) |
|---|---|---|---|---|---|
| Industry Output | Direct | -19,896,420 | -$1,253.0 | -$2,058.0 | -$3,668.5 |
| Indirect | -11,070,952 | -$764.4 | -$1,329.1 | -$2,547.7 | |
| Induced | -17,481,073 | -$982.5 | -$1,688.6 | -$3,019.8 | |
| Total | -48,448,444 | -$3,000.0 | -$5,075.7 | -$9,235.9 | |
| Industry Contribution Analysis, one group | Direct | -19,896,420 | -$1,253.0 | -$2,058.0 | -$3,668.5 |
| Indirect | - | - | - | - | |
| Induced | - | - | - | - | |
| Total | -19,896,420 | -$1,253.0 | -$2,058.0 | -$3,668.5 | |
| Industry Contribution Analysis, separate groups | Direct | -19,896,420 | -$1,253.0 | -$2,058.0 | -$3,668.5 |
| Indirect | -8,002,049 | -$539.1 | -$918.4 | -$1,673.7 | |
| Induced | -13,607,209 | -$756.4 | -$1,304.3 | -$2,324.5 | |
| Total | -41,505,678 | -$2,548.5 | -$4,280.7 | -$7,666.6 |
USING EVENT TEMPLATE
You can follow the same steps as above using any Industry Aggregation scheme. You can even follow these steps and analyze a change across all of the 546 IMPLAN Industries. To speed up this process, consider Using the Event Template to upload all of the Events. Uploading 542 Events (six Industries do not have Output and do not require an Event) will take a few minutes, so don’t panic if you see a spinning wheel or get the message shown here. Just continue to wait a few minutes and they will load.
Once they populate, drag them into your Group on the right and Run. Remember you can use the “Select All” button at the top of the page to highlight every Event.
Running an analysis with 542 Events can take up to 30 minutes. The Results show the same loss in Direct Output, however, the other effects are slightly different, due to Aggregation Bias in the aggregated version above.
NO NEED TO CUSTOMIZE YOUR REGION
In order to run any analyses looking at overall changes like these, there is no need to adjust the Industries by customizing the Region. Any decreases or increases made to Output, Employee Compensation, Proprietor Income, Other Property Income, or Taxes on Production and Imports in Region Details will only slightly alter the Results by adjusting the relationships between Industries in the Customized Region.
RELATED ARTICLES
Written November 24, 2020
Updated July 29, 2021