Payments to tribal government

Michael,

Thank you for your reply.

Another question related setting up our model. We have fairly detailed spending for this project, including payments to subcontactors and other services. I am doing a bill of goods model, using industry change activities for some of the major purchases. My question is... for items that aren't directly related to the construction activity itself (examples: per diem spending by construction workers, payments to tribal entities, procurement spending) should the direct effects that are generated by those activities be counted in the direct effects or should they be included as indirect effects?

Good morning!

With the exception of the per diem spending by construction workers, the other two items should be counted as Direct Effects when recalculating that Direct line. That spending is going to be the worker's local spending on final good's/services, so would not produce any Direct Effect as it is separate from the construction Industry production. For the per diem, you could input the value as Household Income to the region through a Household Income Event (producing only Induced Effects), requiring no further conversion when recalculating the Direct. However, if you want to model exactly what these workers purchased (food, hotel, etc.) you can run each line item through an Industry Event and manually move the Direct Effects from them back to the Indirect Effects when recalculating. 

Best,

Michael Nealy

Thanks, Michael.

We are doing the second method, so we'll run the industry events and then sum the values from the direct effects line with those from the indirect effects line. I appreciate your help!

Monica

Sorry - one more point of clarification. You state that we should move the direct effects "back to the indirect effects" when recalculating. Is that correct? Or would they be moved to the induced effects? They are originating from household spending, so intuitively I would have thought they should be counted as induced effects.

Monica,

By modeling the individual purchases for the construction project, essentially you are manually rebuilding the Industry Spending Pattern. Intuitively, Industries spending on Intermediate Inputs (Industry Spending Pattern) produces Indirect Effects, but with how the analysis is set up you will be getting a portion of you impact as Direct Effects. This is why you need to reclassify those Direct Effects as Indirect Effects when utilizing a bill of goods approach.

Hope this helps!

Michael Nealy 

Actually, now I am more confused. In my first post, I asked about three different types of spending related to a construction project: 

• payments to tribal agencies
• procurement spending (materials)
• per diem spending on the part of construction workers

(note: there are many more spending categories than this in our model, these are just a few examples)

I thought you had indicated that the direct effects from the first two items (payments to tribes and procurement spending) should be counted as direct effects for the total project. Is that correct? Because your last post I am now wondering if I misunderstood. To clarify, we have modeled all of the construction spending categories for which we have budget information as industry change activities, so we are getting direct effects in our results. My intention was to simply use the results from the industry change activities as my results for the construction project, without any manual adjustments.

On the other hand, per diem spending is coming from the construction workers, not from the company. So, to my mind, those seem like they would be induced effects. We do have a good sense for the kind of spending they are doing, so we modeled those impacts using Industry Events, as (I thought) you had suggested. Now I'm wondering how to account for those results relative to the rest of the results I have generated from the construction spending. I would think I would add the direct effects from the per diem spending model to the induced effects, not the indirect.

Sorry if this is confusing. I would be happy to talk over the phone (or zoom) if that's easier.

Hi Monica,

Apologies for the confusion. It sounds like you were given an itemized budget from a client for a construction project and indicated that you wish to use a Bill of Goods approach. What we are not quite certain on is whether you intend to separate out each line item from the budget or just the ones that aren't typically considered part of a construction Industry's spending pattern. Let me explain.

Construction is considered an investment or a series of Final Demand purchases in IMPLAN.

Typically, construction projects are run as an Industry Output Event in one of IMPLAN's 13 Construction Industries (50-62). While it is perfectly fine to use a total construction cost if that is what a client provided, when given a detailed budget there are certain line items which should be excluded from the total Construction Event value (land purchases, FF&E, and other expenditures not typically found in a construction budget like Tribal payments). When those are present, we suggest modeling FF&E purchases and Tribal Payments as additional Events to your Construction Event.

However, it sounds as if you are attempting to model each line item in the budget as a separate Event. In that case, it could get really complicated to determine which is a Direct, Indirect, or Induced effect in your results. Without knowing exactly how you are attempting to model each item, all I can suggest is to use the Indirect & Induced results that IMPLAN generates then "re-create" your Direct Effects outside of IMPLAN using the Construction Budget from the client.

For example, if you were to model the wages for Labor separately as a Labor Income - Employee Compensation Event for $1M, you will not get any Direct Effect for that Event, only Induced. So, when presenting your Results, you would want to add that $1M in as your Direct Effect Labor Income.

A couple things to note:

While FF&E and Tribal Payments are not part of a construction Industry's Spending Pattern, they are still part of the overall construction $ spent. Therefore, if you model those expenditure purchases as Industry Events, you will not need to make any adjustments to the Direct Effects that IMPLAN generates for those Events.

Although you say the per diem spending is coming from workers, they receive that money from someone else – the contractor or investor. If the per diem payments that you mention are NOT also counted as part of the Labor expenditure for either the Construction Event or Labor Income Event, you can model those purchases as well. However, if you are attempting to model the local impacts from construction workers spending their earnings (labor income) while on-site, you must deduct that value from your Labor Income input value used in the Construction or Labor Income Event. Otherwise, you will be double counting! If you choose to model the per diem spending by workers on-site, then you must reclassify the Direct and Indirect Effects that IMPLAN generates for these Events as Induced Effects.

Hope that clears things up a bit!

Angela

Angela,

Thanks for the very detailed response. This is very helpful. Based on your explanations, I believe we have built the model correctly. I have written a summary of our methods. If anything looks amiss to you, please let me know.

1. We have excluded land purchases. The project does not have any FF&E so that is not a consideration.
2. We are using industry change activities for all the budget items.
3. Most of the project is being modeled in sector 56. However, there were numerous budget items related to tribal payments, engineering, legal services, public relations, environmental monitoring, project management, and security (just to name a few) that we modeled in related sectors. This is a huge project (seven years, multi-billion dollars), so it was very difficult to tease out what exactly is “construction” and what is not. We felt the best method would be to use the construction sector as our default but then apply spending to other sectors when it was obvious what the service being purchased was.
4. We excluded non-local purchases (e.g. materials purchased from outside the study area and contracts with vendors outside the study area).
5. We were not given a separate line for wages and benefits (labor income) so we are not modeling a labor income activity. We did, however, make some adjustments to the employment and compensation values in the model based on feedback from the client as to the average weekly wages paid to the construction laborers and related positions.
6. We know that about half of the labor force is non-local so once we had determined the full compensation amount, we reduced it by about half to account for wages that would be leaving the region. Then, we modeled the per diem values separately, as they were not part of the compensation estimates we calculated.
7. Based on your response, I plan to model the per diem spending separately from the rest of the construction and construction-related activities and then reclassify the direct effects as induced effects in the full model results.

Thanks!

Monica

Monica,

Just a couple things to note. 

I suggest reviewing the Spending Pattern for Industry 56 to be sure you aren't modeling any budget items separately that are assumed to be included in the Construction Event. All of IMPLAN's construction Industries include soft cost expenditures such as Architecture & Engineering services. Industry 56 attributes 6.4% of Intermediate Input purchase $ towards that Commodity. If you believe this project spends more than that, then you can of course model it separately, but you might want to consider using an Industry Impact Analysis (IIA) Event Type. This Event Type allows you to modify the spending pattern of that Industry to reallocate the $ spent on certain Commodities and change the Local Purchase Percentage of any Commodity.

I would also suggest to review this support site article on the best approach to adjusting the labor income for known commuting patterns. Employment in IMPLAN is site based, so you would not need to adjust the Employment value. However, IMPLAN models do not assume that all workers live in the same region where they work, so the model automatically deducts Employment Compensation for those workers. If you follow the support site article instructions, you can determine how much Employee Compensation is leaving your study region for non-local workers. For example, if the model assumes that 5% of Employee Compensation is leaked out of your region due to commuters, then you would only want to adjust your input value by 45% not the full 50%. There is a template in the article that you can use to recalculate the effects.

Additionally, if you are modeling the per diem separately, you would want to reclassify the results as outlined (move Direct & Indirect to Induced Effects). However, you may also want to include the initial input values as part of the Direct Labor Income as those are benefits received by workers. For example, if you are modeling $1M in per diem spending I would add $1M to the Adjusted Labor Income Results (as calculated above to account for commuters) of the Construction impacts.

Lastly, as you are modeling impacts that require manual recalculation of effects, make sure to use your Filters in the Results screen to parse out the individual effects by Event. This will ensure that you are not mixing or recalculating the wrong results.

Angela