Retail spending based on Consumer Expenditure Surv

Hello Brit, We aren't entirely certain what you are trying to study here. Rather than providing you with information that may not address what you need, we would appreciate if you could help us understand a little bit more about what you are trying to accomplish with CES data. What overall are you trying to model? Increases in sales to retailers? What the effect of bringing new consumers into the study area would be, etc? Likewise how detailed is your data? Generally speaking you would like to tie the information down to a Sector if possible. Is the data at producer (cost of production) or purchaser (paid at a retail store) price? Any additional information you can provide will help us to get you started. Thanks!

Hello, We are analyzing the effect of bringing new households into an area. I recall that we have done this before by importing household spending patterns, which is probably the most efficient and thorough way to do things. But for consistency with a prior version of this particular analysis, I was planning to enter household spending by various types of goods and services. We estimated spending by category based on the Consumer Expenditure Survey table 2301. So the data is purchaser cost, not cost of production. The categories of spending we are using include: Food (store) Food (away from home) Alcohol Household Operations Housekeeping Household Furnishings Apparel Vehicle Purchase Gas Vehicle Repairs Public Transportation Medical Services Drugs & Medical Supplies Entertainment Personal Care Products Reading Tobacco Does that help? Please let me know if you need more information. Thanks, Britt

Thank you this is very helpful. So it looks like you are trying to split out some of the upper level income categories in IMPLAN. While we can't speak to how the previous study was done, we can offer some options for how to proceed. [ul] [li]Ideally you would want to, where you could, attribute your data back to the producing Sector and then apply Margins to your goods purchases and set [url=http://implan.com/V4/index.php?option=com_multicategories&view=article&id=567:567&Itemid=71#lpp]LPP to SAM Model Value[/url]. However, it looks as if most of your categories may be too broad to do this. [/li][br] [li]You could additionally/alternatively attribute values to appropriate retail Sectors as Gross Retail Sales values, but by doing this you will only see the value of the retailer in the area, and you will lose any possible production, transport and wholesale that could be local. [/li][br] [li]A final alternative, would be to create a custom aggregation scheme for you broad definitions that cannot be easily attributed to one Sector. For these aggregated Sectors you could then apply LPP and Margins. Note that aggregated Sectors are a weighted average of their component Sectors so aggregation does affect the Multipliers and production functions.[/li] [/ul] As regards doing this analysis on a commodity basis or an industry basis, the choice is up to you, but you would want to choose one or the other. [ul] [li]Industry Change analysis allows to designate that a specific industry or group of industries are involved in the production of the desired product. [/li] [li]Commodity Changes, distribute producer based on Market Share values http://implan.com/V4/index.php?option=com_glossary&id=226&Itemid=57 , thus allowing institutions and other industries to produce the commodity. [/li] [/ul]In some instances a commodity is only made by one industry and no institutions, and so the Industry and Commodity Change for these Events is identical. Feel free to respond with additional question.

This is very helpful. Thanks. I do have a couple of followup questions: 1) Would it make sense to model some of the more specific line items (e.g., automobiles, drugs and medical supplies, tobacco) on a commodity basis and the more general items and the services (e.g., vehicle repair, public transportation, entertainment) separately on an industry basis, and then summed the impacts? Or do I really need to pick one or the other? 2) How do I determine which items should be margined and for which items I may need to set the LPP to SAM Model? Thank you! Britt

Hi Britt, Generally speaking we suggest using the industry basis as the analysis since it is more straightforward. However, the IMPLAN Household Spending Patterns are on a commodity basis. In this case you can choose to examine it as you wish, but we would suggest using one or the other simply so that the analysis is performed with a consistent methodology. For many service Sectors and some goods production, the two types of analysis yield identical results (i.e. food service, hotels) because only one industry produces those services. Unless you know that all the purchases are made within the geography of your Model, you will likely want to set all Events to SAM Model Value. You can do this through the Event Options> Change All>Set Local Purchase Percentage> SAM Model Value. Setting SAM Model Value forces the purchases of goods and services by the households you are modeling to match the buying patterns of the region. As regards Margins, you will want to set Margins for all products purchase through a retail environment (e.g. personal care products, groceries, houseplants, clothing and books/magazines). You will not Margin services as these are produced by their seller (e.g. restaurants, haircuts, rental cars, insurance, rent and utilities).