sub-industry analysis

Hi Dick. In reading your Forum Post concerning the impact of street vendors, our spending patterns come from BEA and thus we are tied to the Sectoring scheme and the Intermediate Expenditure patterns they provide. So unfortunately, we don't have any data to break out street vendors of a food service type from restaurants and bars, or street vendors of news/art kind from other vendors in Sector 331. Aside from using the Sectors that you described in your question as a proxy, the other way that you might be able to do this is to edit an Industry Spending Pattern for Sector 331 and/or 413 (if you have a split of sales for both. If you don't and you want to see which one would be a better match your project, the Help>Sector Search sorted by Industry code is a good way to see all the NAICS descriptions included in the Sector. However, to custom build a sector for your purpose, you will need more information about the types of street vendors and their sales in order to edit the spending patterns to reflect your known information, or ideally you can create a custom spending pattern if you know exactly what your street vendors are purchasing. Industry Spending Patterns allow you to import an Industry's production function or build an Industry from data about its expenditures. You can import an Industry Spending Pattern by navigating to Setup Activities> Activity Options> Import> Industry Spending Patterns> Select your Sector to modify. The coefficients listed in an Industry Spending Pattern represent the amount spent on each commodity per dollars’ worth of that Industry's output. An important thing to note is that Industry Spending Patterns typically do not include labor income expenditures, and therefore the coefficients sum to less than 1.00. To ensure that the full impact of spending in an Industry is captured, you will need to create a Labor Income impact to compliment you Industry Spending pattern. The Activity Level can then be used to specify the particular output level being analyzed. Alternatively, if the analyst does not know the total output level but knows the value of input expenditures (that is, the non-payroll portion of the budget), the Industry Spending Pattern can be normalized to sum to 1.00 and the Activity Level set to the non-payroll portion of the budget. Also, it is important to note that these spending patterns do not accounted for direct effect in the results. To get the Labor Income portion of the budget, the analyst could navigate to Explore >Social Accounts> Balance Sheets > Value Added >Sector Search> and enter your sector. From there, you would take the portions of Employee Compensation and Proprietors Income and multiply each value times the non-payroll portion of the budget. Next, you would add these two values back to the non-payroll portion of the budget to get total expenditures for the street vendors. Often user's don't have the ability to specify the spending pattern any more distinctly than we can, but if you do have extra data related to Employment or Employee Compensation and Proprietor Income, these are also editable fields in the Event and modifying this information, if it is available to you, will allow to create some additional specificity for your analysis, over using the regional averages for Output per Worker and Labor Income per worker. We hope this addresses your question. Please let us know if we can be of further help.

Thanks, I am building a model using industry spending pattern. I know “all” of the expenditures in the industry, including employee wages and benefits and proprietor income. The final paragraph in your message is not entirely clear. How would these known data be included in the spending pattern model?

Hi Dick, We apologize about that. You actually won't include any spending associated to labor payments, taxes or profits in the spending pattern. The spending pattern by definition is only goods and services (commodities) that you purchase. Did you set your problem up as an Industry Spending Pattern with coefficients for purchases or as a series of Industry Change Events? If you are using the Industry Spending Pattern methodology, do your coefficients sum to 1.00 or less than 1.00? You will actually model the Labor Income payments using a Labor Income Change Activity. You can enter payments to wage and salary workers (as fully loaded payroll) in 5001 Employee Compensation and payments to proprietors, if any, in 6001 Proprietor Income. Taxes and profits are not re-spent in the Model, so these don't need to be determined unless you are intending to report them. If you know these values exactly, then you won't need to do any computations, if you don't know them there are ways from IMPLAN that you can estimate the split. Taxes and profits will be included in the Direct Value Added and Direct Output values. Hopefully this helps to clarify, please let us know if you need any additional assistance, in setting up your analysis.

No worries, it’s the inherent limitation of this type of communication. I figured that labor would not be in the spending patterns, but would the labor income change activity be part of the spending pattern scenario or is it part of its own scenario? The manual is not entirely clear on that. The main text seems to imply a separate scenario, but then the note to Figure 10-58 indicates that the labor activity should be part of the spending pattern activity. As per advice elsewhere in this forum, I have actually run multiple models. One model is a spending pattern analysis, based on coefficients, with coefficients summing to 1 and using total spending as the scenario level. A second model with a spending pattern analysis, based on coefficients, has coefficients summing to less than 1, with totals sales as the scenario level in scenario 1 and total spending as the scenario level in scenario 2. A third model is built using expenditures as a series of industry change events. So for each model/scenario, how is labor best captured? [quote="IMPLAN Forum" post=17842]Hi Dick, We apologize about that. You actually won't include any spending associated to labor payments, taxes or profits in the spending pattern. The spending pattern by definition is only goods and services (commodities) that you purchase. Did you set your problem up as an Industry Spending Pattern with coefficients for purchases or as a series of Industry Change Events? If you are using the Industry Spending Pattern methodology, do your coefficients sum to 1.00 or less than 1.00? You will actually model the Labor Income payments using a Labor Income Change Activity. You can enter payments to wage and salary workers (as fully loaded payroll) in 5001 Employee Compensation and payments to proprietors, if any, in 6001 Proprietor Income. Taxes and profits are not re-spent in the Model, so these don't need to be determined unless you are intending to report them. If you know these values exactly, then you won't need to do any computations, if you don't know them there are ways from IMPLAN that you can estimate the split. Taxes and profits will be included in the Direct Value Added and Direct Output values. Hopefully this helps to clarify, please let us know if you need any additional assistance, in setting up your analysis.[/quote]

Hello Dick, The Labor Income Change Activity can be incorporated into the same Scenario or run as a separate Scenario. It depends on whether or not you want the Induced impacts summed into the final results or if you want to view them separately and sum them yourself. The book tends to do them separately on a number of occasions for the point of making clear the results of Labor Income portion of the analysis. Here are a couple of questions from the remainder of your response: [quote]with coefficients summing to 1 and using total spending as the scenario level. [/quote] How are you defining total spending here? Total goods and services budget or something else? The Labor Income Change for this Activity as you have it setup would need to be run in a separate Scenario. We presume you are leaving Activity Level at 1.00? If so and you wanted to have the software sum the Induced impacts of the Labor Income Change with the spending pattern results, you could use the Activity Level for budget value and leave Scenario Level at 1.00 and include the Labor Income Change Activity. If you were to add the Labor Income Change into a Scenario with Scenario Level set to budget you would be multiplying your labor budget by that same level. Labor Income will not represent the difference between goods and services budget and total sales. Please let us know if you don't have Labor Income values and need to use the Model to estimate these. [quote]has coefficients summing to less than 1, with totals sales as the scenario level in scenario 1 and total spending as the scenario level in scenario 2. [/quote] Can you provide us some more information about what you are trying to accomplish here, we are little unclear on why you are including the spending pattern twice, are you looking at two different types of Activities? When the Industry Spending Pattern Sum of Event Values is less than 1.00 the Output or total sales value is the correct level. When Sum of Event Values equals 1.00 then you will want to use the commodity goods and services budget value. The same rules about setup for the Labor Income Change apply here, you would need to run it separately if you are using Scenario Level as you sales/budget value. [quote]A third model is built using expenditures as a series of industry change events.[/quote]Are these Capital Expenditures? If so, there will not be a separate Labor Income component to these expenditures. On all these Activities, you will want to be sure your spending patterns have the LPP field set to SAM Model Value. Please also remember to add your Direct impacts back in and adjust the totals for all these impacts for all non-Capital Expenditures spending patterns. Hopefully this helps to answer your questions, please let us know if we can provide any additional assistance.

Thanks, a few follow up questions: When I said I built a model using expenditures as a series of industry change events, you asked: "Are these Capital Expenditures?" No, these are goods and services expenditures. I built a custom sector for vendors by using expenditure data from food and nonfood vendors in industry change analysis. That is, each expenditure type is entered as a new event for food and nonfood respective to the relevant sector. For example, insurance expenditures for food vendors are entered as a new event for code 357 with the annual weighted dollar amount, and so forth. I built a variation on this model by entering all of the expenditure dollar amounts as negative values. The idea here is to examine what the impact would be by the loss of vendors, something cities are doing by adopting regulations to ban street vendors. On a second topic, you said, "On all these Activities, you will want to be sure your spending patterns have the LPP field set to SAM Model Value." I generally understand the 100% LPP and SAM and why one chooses one over another. However, in this specific case, the survey asked vendors where they bought their good and services. Results indicated they bought them all locally. Other than petroleum products (which I set to SAM) would I use the 100% LPP?

Hi Dick, So it sounds like you are building the exact same analysis, but using several different methodologies? If this is the case you will get similar but not the same results between these methods as the Industry Spending Pattern method, is built on a commodity basis and thus will be subject to Institutional production and purchases from inventory, as well as other industries that produce the commodity as a byproduct. The Industry Change method will yield results based solely on the assumption that that industry is affected. You should get relatively similar or identical results to running the positive impacts, because the model is linear. As regards LPP, if you have known values of local purchases, you will certainly want to use these over IMPLAN estimations. However, it is important to keep in mind that sometimes the 'local' purchase is actually a local retailer or wholesaler rather than a local producer. If this is this case, and you know the product that is being purchased, you can Margin the producing Sector and only set the retail/wholesale component. In this case you will want to set LPP to SAM Model value and then in the edit screen you can set the Retail/Wholesale component to 100% local (changing the LPP not the Margin column). If these are the types of purchases you are dealing with, it is easier to do this with the Industry Change methodology than with spending patterns, as the Margining process is 'done for you' in that sense that once you manually set the producing Sector Margins ([url=https://implan.com/v4/index.php?option=com_multicategories&view=article&id=567:567&Itemid=71]Event Options> Edit Event Properties>Margins> Yes[/url]). For the spending pattern method you have to do this process manually as described below: What you will need to do is calculate out the Margin splits for each item that is purchased through a wholesaler, and if you are assuming no retail production, you will also need to adjust the values so that the portion of the Margin assigned to the retailer is adjusted down into the wholesale, transport and production elements of the Margin. So what you will need to do is use the Excel table to match your Sector to the Sector column in the table and then split the value of the purchase across the appropriate listed Sectors. As an example, if you were splitting a purchase of Sector 2 Grain Farming, the table reads: Sector MarginSector MarginValue 2 2 0.388031239 2 319 0.193879404 2 320 0 2 321 0 2 322 0 2 323 0 2 324 0.280330548 2 325 0 2 326 0 2 327 0 2 328 0 2 329 0 2 330 0 2 331 0 2 332 0.003904831 2 333 0.058118416 2 334 0.040410461 2 335 0.0353251 2 337 0 So 38.8% of the expenditure value would go to Sector 2, 19.4% to wholesale (319), 28% to retail grocery (324), ~13.8% to various transport. If you wanted to remove the retailer, you would remove the retailer portion and redistribute or normalize the remaining Margin coefficients. Sector MarginSector MarginValue 2 2 0.539179812 2 319 0.269400631 2 320 2 321 2 322 2 323 2 324 2 325 2 326 2 327 2 328 2 329 2 330 2 331 2 332 0.005425867 2 333 0.080757097 2 334 0.056151419 2 335 0.049085174 2 337 You will then sum the wholesale, and transport components of all the Sectors you know are purchased into a final spending pattern result that will include an aggregate value of wholesale purchases for all commodities. If you are having trouble with the mapping are you using our Help>Sector Search feature? The Sector search can be sorted by clicking the Industry Code column. You can then scroll down through the industry codes until you find the codes you are interested in examining. The associated NAICS descriptions will show you the types of Activities that will report into each category. This is often helpful in determining Sectors. We also have an Excel based version we can point you too if you would like. You will also want to be sure that all your Events in the imported spending pattern have Local Purchase Percentage set to SAM Model Value, this will allow the Model to make estimates of local purchasing ability of all commodities in your spending pattern. [url=http://implan.com/index.php?option=com_kunena&func=view&catid=84&id=17857&Itemid=35]This link [/url]contains the downloadable Margin file. We apologize but we don't seem to be able to attach it here.