Changes in calculating induced impacts?
Hello,
I recently purchased the 2012 state level data for Iowa and I had a quick question for you regarding the induced impact results I am seeing. I entered employment for nine different sectors and am seeing significantly higher impacts from induced effects versus that of indirect effects. I previously ran an analysis using 2008 data that I purchased from you and upon comparing to my new 2012 data, noticed the resulting 2012 [b]induced[/b] impacts are roughly 50% higher than the [b]indirect[/b] whereas in 2008, indirect impacts were 100% HIGHER than induced impacts. I am entering in employment only (in fact only 1,153 jobs in total for 2012) and all four categories I am analyzing (output, employment, labor income and taxes) are quite a bit higher for induced versus indirect. The resulting impact when I enter in roughly the same number of jobs is an increase in overall economic impact of over $140 million. Did you change the way you calculate or create multipliers for induced impacts? If so, when and was this a change in every state and every sector or just selected ones? Can you please help me understand whether it is a calculation change, an error in the data or if there are good explanations like output per worker increased, inflation increased or labor income increased.
As an aside, I just ran a similar analysis for Ohio as of 2012 for the same nine sectors using employment I entered and also noticed a sizable increase in the induced impacts when compared to the same inputs I entered using 2008 data. Although the impact was not nearly as significant as what I described above for Iowa, it came as a surprise to me and seems to illustrate some kind of change in the induced multipliers or underlying calculations.
I would appreciate any help or clarification you can provide ASAP. Thank you in advance.
Kind Regards,
Derek Groff
Was this post helpful?
-
IMPLAN SupportHello Derek, We apologize about the delay. Without knowing which 9 Household Sectors you are looking at, we suspect that: Proprietor Income(and therefore Total Value Added) is a higher proportion of Output in 2012 than it was in 2008. Proprietor income along with Employment Compensation drive the Induced effects. If Value Added is a greater proportion of Output, the intermediate outlays (Output- Value Added) are less per dollar of Output, which would decrease the Indirect effects. Hopefully this helps and if you would like to look into this further, please let us know.0 -
Thank you for your response. Unfortunately, this does not quite help me understand why the induced effects for 2012 seem so out of whack compared to prior years. Below are the nine sectors I have been analyzing for 2012 and comparing to the same ones I used for 2008. I am also attaching a brief file that that illustrate the strange increase in induced effects relative to the indirect ones for Iowa. You will see in the row highlighted that the relationship between indirect to induced effects changes drastically from 2008, despite using the exact same sectors (employment figures entered for the two years are different of course, but not materially so). Also, I am using the Econometric model through multipliers, if that makes a difference. Thanks again for your help in advance. I look forward to your timely response. Sectors: 4 - fruit farming 72 - wineries 113 - printing 319 - wholesale trade 323 - retail stores - building material and garden supply 324 - retail stores - food and beverage 392 - private junior colleges, colleges, universities, etc. 411 - hotels and motels, including casino hotels 413 - Food services and drinking places0
-
IMPLAN SupportHi Derek. We are looking into this issue and trying to see what else might be causing the results that you are experiencing. Please allow us time to try and figure out where in the data is there a problem. As soon as we have an answer, we will share it with you.0
-
IMPLAN SupportWe've done a comparison of the 8 sectors for Iowa 2008 and 2012 (attached spreadsheet). For sectors 113 to 413 there was little change of note. Sector 4 (fruit) has high variation in output per worker and labor income per worker which is not surprising for either a small sector or a farm sector. This is both. Farm prices and productivity can vary widely year to year and region to region. Wineries output per worker and earnings per worker went down noticeably between 2008 and 2012 (more startups/lower wine prices??) which affected the multipliers for that sector. The compensation per salaried employee usually stays consistent but proprietor income per proprietor would go down considerably (if not negative) in either of those scenarios. I would assume that Ohio's wine and fruit sectors would also be the source of the differences. The only other consideration is making sure the comparison between 2008 and 2012 for wholesale and retail sectors are done either using gross margin v. gross sales consistently. However, since you were not entering sales, but rather employment this is tough to get wrong. When you enter employment you automatically get gross margin which is usually what you want.0
-
Thanks for the feedback. It appears you believe the 2012 data is correct and that the primary problem for the extreme change in induced impacts is the fruit farming sector. I am not sure I can agree with that given the extreme change in induced versus indirect from 2008 to 2012, particularly since it went up significantly (and not down as the winery output/earnings comment implies). Finally, you mentioned there was a spreadsheet you attached that did a comparison of the 8 Iowa sectors in 2008 vs 2012. Can you please attach it or send it directly to me? It was not attached to your post. Thanks.0
-
One more comment here and it is important as I have not heard a direct answer back on this. [b][u]Did IMPLAN change the way it calculates Induced Effects for its 2012 data/multipliers? [/u][/b] Also, can you please confirm that the 2012 underlying data used for Iowa is correct? I am having a hard time believing that two sectors (fruit farming and wineries) can account for the majority of what is a $200mm difference between 2008 and 2012 for Iowa when there are only 175 additional jobs between the two of them. These are not high paying jobs or high production/output jobs. Thank you.0
-
IMPLAN SupportHi Derek. We apologize for not attaching the file in our earlier reply to you. Here is the file that was mentioned in the earlier post. Hopefully, when you review it, it will further help to clarify some of your questions. In regards to your later post, we will get an answer back to you before the end of the day. We apologize for the delays.0
-
While I appreciate the response, finally, this file you attached provides no value or insight to me whatsoever. It doesn't answer my questions and provides no help in figuring out why the induced impacts based on 2012 state level data appear to have changed and are inflated in some fashion. I do not have the time to continually ask for an explanation as to why your data doesn't make sense. The continual delays in responding to my questions, and not even addressing them properly, is not only frustrating, but incredibly inefficient and ineffective. If we cannot confidently rely upon your data to supplement our economic analyses, there is little value is using the data going forward. I would appreciate a timely, thorough and final explanation to my original questions and issued experienced with using 2012 IMPLAN state-level data.0
-
IMPLAN SupportHi Derek, There is no change in the way the multipliers are calculated only in the underlying data as noted in the previous spreadsheet. There Is an option of editing the 2012 data (for the fruit and wine sectors) to reflect the productivity relationships of the 2008 data which should give you similar results to the 2008 multipliers. If you send your direct employment for 2008 and 2012, we can use them against the model to see if we uncover any else.0
Please sign in to leave a comment.
Comments
9 comments