Value of Coal and Industry Output

Comments

7 comments

  • Avatar
    DougO
    Yes, output = producer value of production. This is not the same thing as mine sales as some production may go into/out of inventory to satisfy sales. There are two things that lead to the discrepancy you see: 1) We project the value of US production from the previous year to the Implan data year, as producer values are not available at the time of data creation. 2) Much more importantly, we essentially use the average US price to employee compensation ratio and apply that to states. This allows the effort required to mine a short ton of coal to drive the indirect effects, rather than have product quality or location, which increase/decrease prices (and; therefore, profits and output value) also drive the indirect effects. We use cost of employee compensation as a proxy for increased costs of production, but It is quite possible that for local conditions that this does not work. You are welcome to edit the study area to reflect the higher coal pricing, but it will be important to decide how much of that increased valuation goes to profit (increasing other property income) rather than going to increased operational (goods and services purchased) costs. Note, if you do customize the study area data, be sure to uncheck the "Lock" box to allow the reallocation of profits. Note 2, corporate profits (other property income) is treated as a leakage by input-output analysis as we do not guess ahead of time where the shareholders live or how the profits are reinvested. If the analyst has local information and how profits are spent and if that happens within the local economy, then the analyst will need to model that spending as well.
    0
    Comment actions Permalink
  • Avatar
    richardsonlj
    Thanks for the quick response, Doug. A couple of follow-up questions. 1) Does IMPLAN make a *projection* of coal production for 2010 based on 2009 data, or does it simply use the 2009 number? If the former, where can I find that number so I can compare to EIA data? 2) Could you explain this a bit further? Using an average for the US would seem to be problematic, given that the industry is so very different regionally. For example, Wyoming coal is *much* cheaper than Appalachian coal, partly because productivity per worker is so much higher there. Does the model know that productivity levels are different in different states, and if so, how? 3) Your point about updating the pricing information is actually the key to this first part of my analysis. I'm trying to understand how much of the value of coal remains in state, and how much leaves. Corporate profits, it turns out, mostly leave the state, since the vast majority (85%) of production is controlled by just a few companies headquartered out of state. The remaining entities are small and appear to be independently owned, so I'm assuming that IMPLAN's number for proprietor income probably goes to these entities (and I assume that they are held in-state). Does this makes sense? thanks!
    0
    Comment actions Permalink
  • Avatar
    DougO
    The 2009 national number comes from the BEA output estimates from file: GDPbyInd_GO_NAICS.xls which is found on the www.bea.gov website. Our 2010 output number is a projection of the 2009 number. We only have a single production function for the coal industry from the BEA website which would be the National average of all mining technologies and conditions (specifically undergroud or surface for coal). Wyoming average producer price per short ton is indeed much lower per ton (~$12 versus National average of ~$70). Since the supply of coal nationally exceeds demand I would assume that the differences were due to transportation costs to market or quality (eg, sulphur content) of the coal. It is up to the analyst to make the sectors reflect local considerations, if necessary - usually through editing earnings/output per worker and estimated operating surplus.
    0
    Comment actions Permalink
  • Avatar
    richardsonlj
    Hi Doug, I'm still a bit confused about this issue and I wondered if we could explore it a bit further. You mean you are using average U.S. minemouth coal prices? How exactly do you apply this ratio (to employee compensation) to the states? Can you show us the calculation? Since Appalachian coal prices are so much higher than Western coal, using the U.S. average it seems would give a much lower output level for WV. I need to understand how you handle the issue of labor productivity, because I'm actually going to split WV into sub-regions, counties that mine Central App coal and counties that mine Northern App coal. Productivity levels are very different for the two. Average U.S. productivity is projected to decline by 1.4% per year through 2035 (according to EIA), which is about right for NAPP (-1.3%) but not so much for CAPP (-3.9%). I'm not sure how to apply this to the output for both regions, and I can't convince myself that the industry output for 2010 (the default in the model) is even correct. thanks!
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Support
    We do not use U.S. coal prices and production levels per se. Rather, we use U.S. total output (from the BEA). We project BEA Output figures to the current data year using the change in employment from the previous year to the current year (see here for our Employment methodology: http://implan.com/V4/index.php?option=com_multicategories&view=categories&cid=258:employmentdata&Itemid=71). The national output is then distributed to the states and counties based on the national output per worker for that sector multiplied by each state and county’s employment, adjusted for the state or county’s ValueAdded-to-Employment ratio for the sector. For example, if the VA-to-employment ratio for the county or state is 5% higher than the U.S. value, then Output-per-Worker for that county or state will be adjusted 5% upward. Output will then be calculated by multiplying this ratio by the county or state’s Employment figure for that sector. State Output values are forced to sum to the U.S. and county Output values are forced to sum to their state. As Doug mentioned previously, it will be a judgment call as to whether the price increases go to profit or labor income.
    0
    Comment actions Permalink
  • Avatar
    richardsonlj
    Thanks. A few more follow-up questions to see if I've got this straight. For my analysis, there is a significant difference in labor productivity (sector 21, mining coal) in my two study areas, and it’s important to note that labor productivity has been declining in both regions, and at different rates. I’m trying to understand in what context you use national vs. state data. You can see why this is critical in my case, because I need to understand whether (and if so, how) labor productivity changes when you take national output (in a given sector) down to the state and county level. Attached is a screen shot of an Excel spreadsheet. It contains IMPLAN's default Study Area Data (under Customize) for both of my study areas, Northern and Southern West Virginia, as well as the national per worker values for comparison. You can see that the per worker values are the same for intermediate inputs, business taxes, and OPTI. The output per worker is different because labor income per worker (proprietor income plus employee compensation) is different for the two areas. [attachment=343]implan-defaults.pdf[/attachment] Earlier in this thread, Doug says: “2) Much more importantly, we essentially use the average US price to employee compensation ratio and apply that to states. This allows the effort required to mine a short ton of coal to drive the indirect effects, rather than have product quality or location, which increase/decrease prices (and; therefore, profits and output value) also drive the indirect effects.” This seems reasonable—that labor productivity should drive the indirect effects (i.e., how hard is it to get the coal out of the ground). But then, why would the per worker intermediate inputs remain the same for both areas? This seems non-intuitive; if it's harder to mine a ton of coal in one area compared to another, wouldn't the intermediate inputs differ? Later, in the last post, you seem to be saying that IMPLAN takes the national productivity in each sector and multiplies by local employment to put this on the county level: “The national output is then distributed to the states and counties based on the national output per worker for that sector multiplied by each state and county’s employment, adjusted for the state or county’s ValueAdded-to-Employment ratio for the sector. For example, if the VA-to-employment ratio for the county or state is 5% higher than the U.S. value, then Output-per-Worker for that county or state will be adjusted 5% upward.” This last part seems to be how IMPLAN adjusts for differences in labor productivity. That is, by multiplying by the Value Added to Employment ratio (divided by the national number) you apply an adjustment to the national labor productivity number (again, in a given sector). Is that right? If so, where do you get the county’s Value Added and Employment? Are these data from the previous year? Are they static numbers or do they represent the recent trends, which are declining? And what is the source? BEA? Finally, what if I need to change *both* the total output and the output per worker value? In my case, it seems that I might need to change the per worker value to reflect differences in labor productivity, but also change the total output to reflect the industry output for these two study areas. The Customize Study Area Data only allows me to change one, and then it automatically recalculates the other. Many thanks for the clarifications.
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Support
    Our first estimate of state Output is calculated as IE (based on U.S. IE/Employment ratio * state Employment) minus state VA. State VA consists of state-level EC, PI, OPI, and IBT data. After converting to IMPLAN sectoring, state Output is then adjusted so that the sum of states equals the U.S. output value; thus, the state IE/worker ratio may vary somewhat from the U.S. IE/worker ratio. This process is repeated for the counties, but there is no county-level data on OPI or IBT; thus, state-level OPI and IBT data are distributed to the counties based on employment. Thus, all counties will have the same OPI/worker and IBT/worker ratios as the state, and thus the same IE/worker ratio as the state. We have no sub-state indicators to distinguish the intermediate output per workers. The fact that mining is harder in one part of the state is harder than another is your local knowledge and we encourage you to edit the study area data to reflect that information if you wish. The links on this webpage outline our processes for estimating employment and income: http://implan.com/V4/index.php?option=com_multicategories&view=categories&cid=241:datainformation&Itemid=71 If you change Output, the Output per Worker automatically recalculates to adjust to your new Output figure. If this does not give you the desired Output per Worker, then you need to also change the Employment figure. Jenny MIG, Inc.
    0
    Comment actions Permalink

Please sign in to leave a comment.