Wholesaler Impact

Hi Bart, If we can we would like to back up a little and ask what value you are actually working with. Are you looking at the value of sales to consumers of craft beer, or are you looking at the sales value of the producer's production, or the value of wholesales sales of craft beer. These values will be handled in very different ways in IMPLAN depending on what your original value actually represents. As regards what you have done the Gross Retail Sales is what you use when you are working with the sales receipt value of items sold retail from a store. Gross Retail Margin is for the operational value of the store, i.e. the electricity, employee costs, janitorial costs, waste management costs, etc of the retailer. Neither of these methods will actually show you wholesaler values. IMPLAN does have a wholesaler Sector, but this is very broad and will again show you only the value of the wholesalers operations. Please let us know what value you are starting with and we can hopefully point you to something that would better meet your needs.

I'm trying to calculate craft beer's share of the value created by beer wholesalers. I have data across the value chain (value of craft breweries' sales to wholesales, wholesalers to retailers, and retailers to consumers) and so can also calculate the value added at each step. I'm trying to put something through the wholesaler sector in IMPLAN to come up with an indirect/induced share of wholesaler value added. My questions is: when I do this based on gross retail sales or gross margins (for the wholesalers) I get quite different values. My hunch is that this is because (as you say) the wholesale sector is broad, and so it's assuming an average margin on the final retail value that is quite different than the actual margin in beer. In that case, should I be entering the value as wholesaler margins?

As a follow up question... if I enter both a wholesale and retail sector into a model, is the indirect impact from the retail sector double counting the same impact that was entered directly in through the wholesale channel?

Hi Bart, If you have data all the way across the supply chain, that's great. In this case you use the advanced Margining method and then edit the Margin percentages to be exactly what you know. This method will avoid a lot of extra work and/or double-counting concerns. In regards to the Gross Retail Sales and Gross Retail Margin concerns, we will try to explain it more clearly. As you know from your value chain, a retail purchase doesn't represent the amount of the retailer's production but represents the cost of the entire value chain. So if I buy a 6-pack of beer at the grocery store and pay $13.00, let's say that the retailer will keep $3.00, the wholesaler may get $3.00, the transportation chains will get ~ $1.50 and the brewers gets $5.50. Because of this when I make a purchase at a retail store even though I pay $13.00, the retailer only gets $3.00 for their operations and employee's and profits. When you are working with retail and wholesale Sectors (319-331)the Gross Retail Sales is the default because typically the value people bring to the model is the receipt cost, $13.00 in our example. Thus because a variety of products can be purchased at a food and beverage seller, and some may be locally produced and other not, the only value kept in this circumstance is the $3.00 that the retailer keeps from the $13.00 sale. The $3.00 is then spent on the items the retailer purchases for it's operations, utilities, payroll, etc, but not the products sold through the retail outlet, because it doesn't produce the products, the retailer produces the services of selling all the products made by a variety of producers in a single location for the convenience of the final consumer. You get a much larger result, although incorrect, if you enter the $13.00 as Gross Retail Margin because you are telling the model that you have already removed the $10.00 that go to the wholesaler, transport and producer and you are just providing the value of the retail operations. The money is still spent ONLY on the retailer's requirements for operating the store. So from what you've described is a way to look at the whole value chain, and this is something we can do with the more advanced form of Margining. Typically we recommend using the retail store only when you don't know what the specific product being purchased is, but when you know who the producer is, as you do in your case, you can start from the producer [i]71 Breweries[/i] and then manually apply Margins Event Options> Edit Event Properties>Margins> Yes. Since you have all the value-chain information you will then want to edit the Margins to match the information that you know via Event Options> Edit Event Properties>Margins> Edit. This will open up a screen where you can shift the values of the Margins to match the proportions that you know meet your particular industry's measures. Depending on how you are looking at you study, and the size of your region, you may want to assume, or your numbers may be restricted to reflect only values of transport, wholesale, production and retail that are local to your Model geography. If this is the case you will want to leave the Local Purchase Percentage at the default value of 100%. If however, your sales values are not limited to the Model geography, you can set the Local Purchase Percentage equal to SAM Model Value via Event Options> Edit Event Properties>Set Local Purchase Percentage> SAM Model Value pathway. This will allow the transport, wholesale, production and retail to be purchased according to the Model's values of local use of local supply. You can also custom edit the Local Purchase Percentage values, if you have that information as well in the Margin>Edit screen. With this method, you will want the Industry Sales value to represent the total ticket cost of the retail sale of your craft beer, the customize Margin split that you edit in the Margin> Edit screen will then break that sales value down into the respective contributing Sectors and the Model results will report each element as a Direct effect, and then show you the resulting stemming from effects. This method will combine Indirect and Induced impacts along the entire value chain for you, so that you can demonstrate the full impact of your industry. Please let us know if you have any additional questions.

Thanks for the great breakdown. Hopefully final follow up question. Is there a way to add sectors in the edit event properties>margins window? I ask because 71 Breweries only has off-premise retailers, and doesn't allow me to allocate margins to bars/restaurants.

Hi Bart, That is a great question and we are trying to come up with the best answer for you. The problem lies in the fact that restaurants are not retailers, in that they actually produce the food they serve, or modify it in someway, so they do not behave the same way in the model. Do you have a split between the proportions of beer sales associated to retail stores and those to restaurants and bars for the wholesale, transport and production as well, or just a split of final sales value? We presume that you are just trying to capture the impact of the beer sales, not the remaining restaurant activities, is that also correct? Thanks for the additional information.

It's actually one level more complicated than that. I'm trying to come with just the impact of craft beer sales for normal restaurants, but beer + other activities for brewpubs (restaurants that also produce their own beer). I'm solving that currently by just pulling out the brewpub sector and treating it separately. As for the split: yes, I have both sales volume and final sales value. My biggest worry with the margin approach is that by sending it through food retail I'm way underestimating employment, since a much larger chunk of craft beer value goes through on-premise (bars/restaurants) and those have higher employment relative to sales.

In this case, would it just be better to send the retail sales through the separate sectors: food/beverage retail + bars/restaurants. I guess this returns the worry about double counting value, but eliminates the problem with misestimating employment/wages. In that case I'd have: 1. Breweries (enter just their revenue) 2. Brewpubs (enter total sales) 3. Food beverage retailers (enter their share of sales as gross retail sales) 4. Bars/restaurants (enter their share of sales as gross retail sales) 5. Wholesalers (in this case, better to enter as gross retail sales 3+4, or as gross retail margins [wholesaler margins] -- I have both). Does that make sense?

Hi Bart, It certainly does make sense. So here are our thoughts. In regard to the Brewpubs, the two functions that they are preforming will in fact behave much like that of two separate entities, except that there will obviously be no transportation and no wholesale factors to this Direct production to sales method. We presume that in either circumstance, Brewpubs or restaurants that you sales value represents only beer sales and does not represent the remaining food or other sales. So that in the end what we are looking at is basically a service Margin for serving beer and it's production, which will not be significantly different from general Breweries. The key thing that you would want to take into consideration is actually if there is a significant difference in wage rates between 413 and 324 in your region. Employment actually doesn't drive any impacts so it won't affect how we run the analysis. In order to capture the Employment difference you can just split your sales value or restaurants and retail beer sales and estimate Employment from the Study Area Data's Output per Worker ratios, found in the Industry Summary sheet in the View By: drop-down window. We would still recommend using our original breakdown Margining Sector 71 and adjusting the Margin values to meet your known, as the service associated to the serving beer in pub/restaurant will not really be different to the service of selling beer in a retail store. You can then use your known splits to divide the retail impact between the restaurant/pub sales and the store sales. If you are still concerned about the Brewpub, you could setup two different Activities following the method of Margining 71 one of non-Brewpub Scenarios editing the Margins for the wholesales, transport, retail and production and then set up a separate one for Brewpubs, to 71 where you zero out the wholesale and transport margins and just allocate the sales and production based on your splits for these two items. Again in this circumstance you can edit IMPLAN's projected Employment value with the ratio for 413 if you feel that this is more appropriate. One other thought on that though, the number of people it takes to run a kitchen and wait staff, while larger, may also not be representative of an environment associated to the service of pouring beer for customers. The Employment rates of the retail store may actually be more reflective. You certainly have more industry specific knowledge on this though, so that will be your call. Please let us know if you need any additional assistance, or if something we have said isn't clear.

Wonderful. Two small questions. How do I use the "known splits to divide the retail impact between the restaurant/pub sales and the store sales"? Is that simply done after running the model? Secondly, since wages draw indirect and induced impacts, is there a way to account for the difference in total wages that will occur once I split the retail impact?

Oh, and for the wholesalers, better to enter as gross margins, since these will be different in beer wholesaling than in the larger wholesale sector (which is what the model will assume if I enter gross retail sales)?

Hi Bart, You should just have 1 Event for Sector 71 with Margins applied and then you divide the Margin percentages in the Edit screen to reflect your relative proportions of the total sales, based on what we have described ( or 2 Events if you are splitting the Brewpubs from the remaining retail/ restaurants) If you are building each of these as separate Events (i.e. an Event for Sector 71, and Event for Sector 335, and Event for 324, you do risk some double counting). In regards to the splits, yes you can do this portion with the Direct results of the retail portion and the associated retail Employment, in affect just dismissing the Models projection and replacing it with you calculation. In regards to the labor difference the easiest way to handle this would be to determine, using the Labor Income per worker values for the two industries, found in the same summary sheet, and your Employment in each subset to determine whether the income values will be higher or lower than the projected income in the model (which can be seen in the Detail Results> View By: Labor Income). You can then run a Labor Income Change Activity type for the net value (positive or negative depending on if the retail Sector's Labor Income per worker is higher or lower than restaurants). You can run this in the same Scenario, so that the software will add this difference into your reported results. Please let us know if you have any additional questions. IMPLAN Support Team

Great - thanks - hadn't thought of the labor income change.

Next question. The margins approach works great at the national level, but how should I make this work at the state level? My guess is I will have to go in and adjust the local purchase percentage, but would love thoughts on how to do this. The major problem is that there are states that are major exporters of craft beer, so that brewery revenue + wholesaler margins + retailer margins are much larger than the total sales in the state. In that case (or the opposite, where retail sales exceed the three things above), do I just need to adjust local purchasing percentages? What should those be set to?

Hi Brian, The model will not restrict in regards to whether or not the sales are larger than the production, or production larger than the local sales. If you have the value-chain breakdown at the state level as you did in your national study you will still want to edit this value to represent you know about each step in the value-chain; if you don't have this breakdown, you can use IMPLAN's estimations but because we only have the Margin breakdown at the national level the splits between retail, wholesale, transport and production will remain static between states. The Local Purchase Percentage, on the other hand, can be set to reflect the value of local supply to meet demand, so if you have the splits but you don't know how much of each of the Margined portion is derived from the local economy, IMPLAN can estimate this for you. These will of course be somewhat broader than just those related to craft beer. So the LPP setting will reflect the average use of locally produced Breweries, various transport for all commodities for that transport type, local rate wholesaler for all wholesaled products local rate for food and beverage purchases for those types of retailers. To let the model make estimations for you, you can set the LLP to SAM Model Value through the Event Options>Edit Event Properties>LLP>SAM Model Value. When you make this change to the Margined Sector 71 Event, it will apply these values across the entire Margined value chain, which you can verify by examining the Event Options>Edit Event Properties>Margins>Edit. If you are fortunate enough to know one or all of the local use of local supply percentages for your sales, the Margins edit screen is also where you can make these changes. You will notice that the screen contains a column at the far-right that shows the LPP values. If you place the cursor into these fields and type in you percentage, as a decimal, you can change the estimations to your known values. Thus if you know all the retail sales were local, you can adjust the retailer to 1.00 or 100% and let the model estimate the remaining values. Please let us know if we can help further.

Does the model account for beer excise (and other sales) taxes in the Brewery (sector 71) path? I'm currently taking them out of sales so that the output is purely based on the value of the sale of goods (and not the taxes) but realized that it might.

Hello Bart, The IMPLAN Output for Sector 71 will include those taxes, so you definitely want to include them in his 'Industry Sales' value. If you don't feel that the Direct portion of the Tax Impact report correctly reflects what you know (and thus also the Total tax impact will be affected if you change any of the Direct values) then you can modify the reported taxes to match what you know. Federal excise taxes will show up in the "Indirect Bus Tax: Excise Taxes" row of the Tax Impact Report. However, as can be seen in the attached Word document, the Census of Government Finances (where we get the State/Local government data) does not have a separate code for excise taxes. Our guess is that they are therefore,lumped in with sales taxes. [attachment=445]TaxImpactReport_2013-10-08.pdf[/attachment] [attachment=446]2011SLIndivUnitDataFileTechDoc.doc[/attachment] Please let us know if you have any additional questions.

One similar follow up. If I'm using a margins approach, should I also include the value of the taxes in the margins assigned to each sector, or should those be taken out?

Hi Bart, Excellent question! The taxes should be included. Each aspect of the value-chain will have it's own Value Added and one part of this is Taxes on Production. Additionally since Output = Intermediate Expenditures + Value Added, the Value Added is a portion of the Output (Industry Sales) value. So you will definitely want to include it. If you known values differ from IMPLAN's estimations you can just modify the Direct Tax impact from the model to match what you know and then re-sum the totals. Please let us know if you have any additional questions. IMPLAN Support Team

Any suggestions about what to do in a case where the revenues of craft brewers are larger than the size of the retail market for craft beer in a state? I've been using the margins approach to great success, but in one state, there is a relatively small craft market and they have one very large exporter, so brewer revenue actually exceeds the retail value of craft beer that is sold in the state. Doesn't appear that I can set margins above 100% Thoughts? I can always enter the channels separately, but then will have the aforementioned problems with double counting.

Hi Bart, You are correct that Margins cannot be more than 100%. If we recall correctly, you have the exact value of sales for each portion of the market chain that is local, and you are adjusting the margins for each component to represent your known proportions. Since you know you are looking only a sales local to each geography, you have at least retail or perhaps all Margin components set with Local Purchase Percentage = 100%. If this is correct, then we would recommend following the same initial process. You will enter the full retail sales value into Sector 71 and apply Margins. You can then adjust the Margins for the retail, wholesale and transport components to their respective proportions based on the total retail sales value and let the producer (breweries) just accept the balance. Next set the Local Purchase Percentage for Sector 71 to "0%". This allows the Margined portions to be captured but "leaks" the breweries impact. Next you can create a second Event for Sector 71 that includes the value produced by that Sector locally. (If this is a production value, then you can just use that value as Industry Sales, if there is any retail component than you will want to adjust the value by the producer margin for Sector 71). This Sector you won't apply Margins to because the Industry Sales value will represent just the production component. Hopefully we didn't make that too convoluted, and please let us know if made a wrong assumption or if our answer isn't clear. Thanks!

Makes perfect sense. One follow-up, in the second event for Sector 71, do I need to subtract the portion that was margined in the first event, or will setting the LPP to 0% negate all impact there?

Also, Should I set employees/wages to 0 in the margined sector and include all brewery employees/wages in the 2nd event, or will the 0% LPP again take care of that?

Hi Bart, Setting the LPP for the Margined portion of breweries will leak all of the assigned dollar values, so you shouldn't need to worry about subtracting anything or adjusting anything. You'll be able to verify this if you run the impacts in the same Scenario by looking at the Detail Results View By: Output tab. You will see that only the value you entered into the second unmargined Event will be reported for Direct Sector 71 Breweries sales. Please let me know if you need any additional assistance or have additional concerns.

Great - Thanks!