Which industries produce a commodity?

Comments

15 comments

  • Avatar
    jenny
    Explore > Social Accounts > Balance Sheets tab > View By: Commodity Balance Sheet > Industry-Institutional Production tab.
  • Avatar
    ndennis
    In my Siskiyou County, CA model, why is Institution 11001 (federal government non-defense) credited with producing $1.1MM of commodity 3015 (forest, timber, and forest nursery products)on the commodity balance sheet, while Industry 429 (other federal government enterprises)is not credited with producing any commodity 3015 on the industry balance sheet? Are production levels of federal government institutions and industries somehow not equated or balanced?
  • Avatar
    jenny
    Institution 11001 is distinct from sector 429. Sector 429 includes things such as government-owned and operated liquor stores, airports, sewer and sanitation services, gas and water supply, and post exchanges (which provide merchandise and services to military families). See page 239 of this user's manual for a complete list of activities included in sector 429 (which was sector 496 in the old 509 sectoring scheme): http://implan.com/v4/index.php?option=com_docman&task=doc_download&gid=105&Itemid=7. See this webpage for a description of these special sectors: http://implan.com/v4/index.php?option=com_content&view=article&id=688%3Aspecial-sector-definitions&catid=187%3Afaq&Itemid=1
  • Avatar
    ndennis
    Thanks, Jenny. With a name like "other federal government enterprises", it would seem that industry 429 would include all federal commodity production not accounted by the other federal government industries, i.e., 427 or 428. If none of 427, 428, or 429 produce commodity #3015, which federal government industry does? I understand the Institution 11001 is distinct from Industry 429, but doesn't the federal institution's commodity production have to show up in some federal industry (that's what I meant by institutions and industries being "somehow balanced or equated")?
  • Avatar
    jenny
    Government enterprises (i.e, those government activities that are treated as industries in IMPLAN) are government agencies that cover a substantial portion of their operating costs by selling goods and services to the public. They operate much like private sector firms, hiring labor and purchasing other inputs to produce goods that are sold through markets. By contrast, government institutions (i.e., administrative government activities) are not an industry but a final demand. As such, they do not have "output" per se. While it seems counter-intuitive that the forest service would be a government institution rathter than a government enterprise, it falls under the umbrella of the USDA, which is a largely administrative government activity. If you would like to see the impact of federal timber production, you can use the private sector (sector 15) as a proxy, customizing the Event so that PI and OPI = 0. Note, however, that the majority of the impact from timber production occurs downstream - that is, in the logging and sawmill sectors - which will not be captured by an impact to sector 15. If you want to see the impact on these sectors, you'll need to somehow translate the value of timber production into some value of impact to these downstream sectors.
  • Avatar
    ndennis
    Thanks: Now we're getting somewhere. I think you're saying that the $1.1 MM in commodity 3015 production credited to Institution 11001 for Siskiyou County does not show up as production by any federal government sector. Correct? I do need to analyze the impacts of federal timber sales, so I will take your advice and analyze them using an event involving production by industry 15. What are PI and OPI?
  • Avatar
    jenny
    That is correct. PI stands for Proprietor Income, which consists of payments received by self-employed individuals and unincorporated business owners. In the case of the government activity, there will not be any PI. OPI stands for Other Property Income. It includes corporate profits, capital consumption allowance (CCA), payments for rent, dividends, royalties and interest income. OPI is treated as a leakage in impact analysis anyway (it is not part applied to the multipliers and will have no additional effect) so it won't affect your results whether you zero it out or not - we just typically recommend doing so when using a private sector as a proxy for a government institution only so that the direct OPI is zero since government institutions are non-profit entities and it would be difficult to estimate their OPI based on the private sector's OPI. You will also want to zero out the direct IBT (Indirect Business Taxes) in your tax impact report, since government institutions do not pay taxes (there is no way to zero out IBT during the impact set-up).
  • Avatar
    ndennis
    Regarding capturing the downstream effects from sector 15's output (e.g., logging and sawmilling), my practice is to model an event based on the output of the furthest-downstream industry located in the region, which, in Siskiyou, would be veneer production. For logs or stumpage that is exported, I model events based on output for those sectors. Is this how you would do it? Back to government institutions (e.g., the Forest Service): I understand that, since their activity is treated as final demand, rather than output, it has no downstream effects, unless modeled using a proxy such as sector 15. I presume, however, that Forest Service activity does have upstream effects (i.e., multipliers > 1), based on their purchases from the region. Is this correct? I'm thinking here about the indirect effects of activities such as stewardship contracting, i.e., purchases of local services.
  • Avatar
    jenny
    If you can assume that the veneer mill relies on the local supply (and wouldn't import) then yes, model the most forward link to get the impact. The veneer mill as part of its impact will capture the backward links of logging and timber production. You would model all end users of the stumpage whether it is veneer, export or pulp. Note, you would not export stumps, but rather, logging. Stumpage is only exported by purchases of outside loggers. The government institutions have no multipliers associated with their activity. You would have to model either through the Non-defense government institution (Setup Activities > Activity Options > Import > Institition Spending Pattern) where the US Forest Service would be incorporated, or by using the private sector 15 as a proxy for the FS activities.
  • Avatar
    ndennis
    1. I agree with you that the size of the event occurring at the most-forward-linked sector must be adjusted to take into account upstream leakages. For example, if the local national forest increased timber sales by 5 MMBF, with half of this increase being exported and half going to local veneer mills, the size of the event affecting veneer mills (sector 96) would be proportional to a 2.5 MMBF increase in log processing. 2. Your statement "Stumpage is only exported by purchases of outside loggers" raises an IMPLAN problem. That's not how it works here in northern California. Loggers generally sell logging services (including log hauling) to landowners; they don't purchase or sell logs. Timberland owners (sector 15) either sell stumpage (3015) or delivered logs (3016). If they sell stumpage, it is often to mills outside the county, so sector 15 does in fact export stumpage. The industry balance sheet for sector 16 (logging) shows that loggers only sell one commodity: logs (3016). They are not shown as selling logging services. In fact, there is no output similar to logging services listed in the entire Sector Search database. How would you suggest I model local logging industry activity? Should I adopt the IMPLAN convention of assuming loggers buy stumpage and sell delivered logs? 3. The commodity balance sheet for 3015 (stumpage) shows that sector 10 (all other crop farming) dominates stumpage sales, and that timberland owners (15) are a minor source of stumpage. This is inaccurate. I presume sector 14 approximates the ranching or nonindustrial timberland owner sector, which does sell a small amount of stumpage. However, stumapage sales here are dominated by industrial landonwners (15). This raises the general question of how to best edit the balance sheets and other databases to correct these types of errors. The main objective of the project I'm working on is to validate the IMPLAN databases to make the model as defensible as possible. So far, I've found about a half dozen major database discrepancies I'd like to edit. What's the best way to do this? Best regards-- Nick
  • Avatar
    DougO
    2. You are right, it was (past tense) a BEA I-O definitional thing that stumpage is only sold to logging, but that hasn't been true since the 1982 benchmark. Looking at the commodity balance sheet, stumpage appears to be mainly sold to logging and sawmills and some to veneer. Curiously, none is sold to the pulp sector. However, even if the stumpage is exported, you will want you use the logging sector as that activity will still be necessary to export the stumpage. My suggestion is to specify the final local use of the stumpage (veneer, sawmill) as well as the total logging and stumpage production for the impacts. You will need to set the RSC's to 0 for those commodities (stumpage, logging, sawmill, veneer) so those activities are not double counted in the indirect effects. Note that port activity is a forward link to logging for exports so you will want to model that activity as well for the foreign exports if it is important. 3. This is not true at the National level and I suspect sector 10 mostly produces x-mas trees. Private stumpage output is difficult to estimate as many landowners to not have wage and salary employees, but rather use, forestry services. It is employment and income in the stumpage NAICs that is used to distribute National output estimates, so regions relying on forestry services get shortchanged. If you follow the suggestion in 2 above, you will still be able to specify all stumpage production as the event as the multipliers are based on the per dollar output.
  • Avatar
    ndennis
    Doug--I've got good data from a state agency on timber sold (both volume and value) in the county by the private and public sectors, so your statement that "private stumpage output is difficult to estimate" isn't true. I've also got reasonably good data from state agencies on employment in each NAIC sector from 0 to 6 digits. I'm planning to use this independent data to revise the IMPLAN balance sheet databases, which in several cases are obviously "way off". As I said in my last message, validating (i.e., correcting) the IMPLAN databases is the main reason the county hired me for the project I'm working on. My question to you is what is the best way to edit the IMPLAN databases?
  • Avatar
    DougO
    While it isn't true for Oregon, we would have to have to collect consistent data for all states/counties in the US from each. It is impossible to even get data like the QCEW (ES202) data from all states on a consistent basis (most states do not subject state and local government employment and income to non-disclosure rules - while the national data source (BLS) does. Most states would not give the breakout between ownership types). This is why we recommend the users customize data with local sources if it is important to the analysis. Also, note that sector 15 is more than just stumpage (but in Oregon I imagine it is most of it). When editing, remember that the employment data includes proprietors/land owners and proprietor income which is important for stumpage and logging (sawmills and veneer not so much). In sector 15 proprietors greatly outnumber wage and salary workers - similar to farming. There is a document on the IMPLAN website which you can use to convert the wage and salary data to employee compensation. It is the labor income that drives the induced effect so it is important to get that correct (employment is "merely" descriptive unless you are estimating output based on output per worker). If you are customizing just forestry related sectors it is probably easiest through the software (Customize>Study Area Data).
  • Avatar
    ndennis
    "There is a document on the IMPLAN website which you can use to convert the wage and salary data to employee compensation. " Which document? How do I find it?

Please sign in to leave a comment.