You are correct in thinking that the multipliers are linear and all changes should be proportional - all other things being equal. Just have to figure out what is not equal. 1. If you are applying an industry change to sector 38, the only way I can think of getting a "net leakage" is to set the LPP to something other than 100%. How are you calculating net leakage? 2. In the first run, did you customize employee compensation or proprietor income and not customize it in the second run? 3. Are you setting the year of impact before entering Value of Sales and using consistant year of event and impact results in the comparison?
I calculate "net leakage" as Total Direct Spending (Emp Comp + Prop Inc + Sum of Direct Regional Inputs) - Indirect & Induced Output. For my two runs: 1. \$16.4 million + \$6.8 million - \$8 million -\$8.9 million = \$6.3 million net leakage. 2. \$16.6 million + \$6.9 million - \$8.1 million - \$9.9 million = \$5.4 million net leakage. Is that appropriate? I did not change the #38 Institution at all for either run. I am letting the model keep 2002 as the year of analysis for both cases. Christy Risch
Curious definition of "net leakage". The 16.4 and 16.6 millions is the value of construction, and that value includes payment to labor (6.8 and 6.9 millions) so it looks as though you are double counting that piece. It may be more intuitive to look at the industry balance sheet for industry 38 to see the gross outlays for goods and services and the amount paid for local goods and services. The difference being the imported goods or services which represents leakages of the first round of spending by the construction sector. The "leakage coefficient" could then be applied to your original direct effect to see import leakage. The 8.9 vrs. 9.9 induced is not proportional. An early version of V3 would cumulate induced effects, so make sure you have the latest version (under Help>Check for Updates). The current version is 3.0.9.2. If that is not the problem, what year data are you using and which county file? I would like to try it myself.
I should have been more clear. the \$6.8 and \$6.9 million are from the balance sheet. They are the sum of amount paid for local goods and services. the \$16.4 and \$16.6 million are the direct income amounts. The total construction project values are, 1. \$33.9 million and 2. \$34.4 million. I'm trying to get a total leaking including the indirect and induced leakages. Is this an incorrect way (or unconventional way) to calculate this? My county is Kanawha County, WV. It is a pretty diverse county but I want to make sure everyone understands how much gets leaks out for such a small project in such as small area. We have version 2.0.1025 - we are way behind evidently.
Version 2.0.1025 is the last of the version 2. Which year data? Your method will not derive a "net leakage". Let's say the economy is completely closed - no leakage at all, a direct effect of \$1 million will have infinite indirect and induced. Or even if there is a tiny amount of leakage, the indirect and induced will be extremely large relative to the direct (think of what that will do to your equation). Now whether there is a tiny leakage or the entire \$1 million leaks out the first round, the "net leakage" is 1 million dollars. The \$1 million circulates through the economy until it all leaks out (taxes, savings and imports). So if the total construction project was \$33.9 million, the "net leakage" will be \$33.9 million. What is important is that it caused additional amounts of local indirect and induced effects before it disappeared from circulation. The slower it leaks out, the more indirect and induced effects it will cause. The 16.4 or 16.6 labor income may have been the original amounts to go to labor, but the induced effects capture the circulation of that money through the local economy before it leaked away.
The income effects, like the employment effects, are easy to look at as you just add the indirect and induced to the direct. I have a lot of trouble explaining the output effects, so please help. This is because the project value/direct amount is not all spent as a large share is immediately spent outside the region (for this case about one-third) and then the same thing happends with the first true round of indirect spending. So what is the best way to describe the economic impact to output? Value added? I came up with the "net leakage" method to describe what happens to output beginning with that initial large leakage. The final amount usually ends up being close to the value-added amount, which made sense to me.
Output and employment have the same implications. It is much easier to see if all industries have the same output per worker. Let's assume that the output per worker for all industries is \$100 thousand. If there is a demand for a new \$30 million structure: Direct Output = 30 million Direct Employment = 300 Indirect Output = 10 million Direct Employment = 100 Induced Output = 15 million Induced Employment = 150 Much of the initial 30 million is spent on imports, so the indirect total is smaller than the 30 million just as the indirect employment is less than the original direct. The entire 30 million is spent, it just isn't all spent locally, but there is demand for that 30 million structure and that is the direct effect. There is still something odd about your induced effects - what year data were you using?
Thanks for the explanation. The relationship between direct output and employment is clear. What is less clear is the relationship between direct output and local output. I just checked and it was the 2002 data. We have 2006 data but I neglected to call up that folder. I will rerun it with 2006 data.
Direct Output = Local Production (unless you have set Local% to less than 100%). That 30 million dollar structure was locally produced. The goods and services that went into its production (first round indirect) do not have to be local. The concept is no different than producing \$30 million worth of cars for export. The production is local even though the goods and services that went into them may have been imported. (You can't export/import a structure but it is still local output.)
I see what you mean. But the whole \$30 million is not part of GDP right? Only about \$16.5 million is value added, which equates to GDP, correct?
True, GDP = Value added. Value Added = Output minus intermediate outlay. Using 2002 Kanawha, I get 8.87 induced from 33.9 direct and 9.00 induced from 34.4 direct (millions).
I got 9.9 induced with the last 2002 run. Using 2006 data I get \$10 million induced and only \$6.4 indirect. Ouch, that was a bad 4 years. The employment impact is a lot lower, 493 vs 642 originally. On value-added I know it does not support what I was trying to do because neither local or non-local intermediate purchases count.
I want to bother you with one more question on this concept. For this scenario is it correct to say that "IMPLAN estimates that the \$33.9 million project creates \$16.2 million in output/sales in the local economy during the one year project (the amount of indirect and induced output)?
Usually, output per worker (lower direct employment) increases over time, especially in nominal dollars. If you tell both the 2002 and the 2006 models that the 33.9 million is 2011 dollars they will be more consistent with each other (but lower direct yet).
I would say it as: the project creates an additional \$16.2 million of new local economic activity. (You can replace "activity" with "production" or "output" and maintain the same meaning.)
But this is not in addition to the \$33.9 million because that would be double-counting the first round of indirects. This is in addition to the status guo, right?
Actually, no, the Building worth 33.9 is an economic output. There was a local demand for production of this value. The 16.2 million of new sales were the local sales that resulted from that initial demand. If were possible for the local contractor to import all the materials and all the labor who managed not to spend any of the labor income locally, than conceivably there would be no indirect and induced associated with the 33.9 million direct effect.
Mathematically does the indirect category include the first round of direct spending?
Correct.