Comments

25 comments

  • Avatar
    ScottL
    Using the trade data in that manner is fine, but that is not a standard supported use of the the IMPLAN system at this time. The trade data is there for the purposes of MRIO impact analysis. I'll have to look at the 2007 data. There were changes in the routines between data sets refining the process but I will have to ask Doug and he is out until next Tuesday. 1. they are in millions. 2. The industry commodity production information is in the CommodityBalances table. 3. You have to calculate that. 4. I'll have to take a look.
  • Avatar
    ScottL
    What state are those in?
  • Avatar
    ScottL
    I figured it was Colorado. The main difference is in the construction flows. I will have to show this to Doug. He will be back Tuesday. I know there was an issue with this in a couple of states for 2009, but I thought it was fixed awhile ago.
  • Avatar
    ScottL
    BTW, there is a lot of documentation on line, but as I said, the documentation is all directed at impact analysis.
  • Avatar
    rochette
    Thanks Scott. Yes, it is in Colorado. We had a non-traditional dataset for 2007, as it was specifically created by Implan to have trade flows for inner and inter regional analysis, as well as seeing things from one county to the rest of the country, all through Access queries. 2,3 - it looks like the domestic (within country, but out of county?) data and foreign (out of country) data are all in the RegionalCommodityBalances table. Can you assist me is seeing which columns to use for dollars transferred? My guess is "ForeignExports" for the model-region-to-rest-of-world numbers, "DomesticExports" for the model-region-to-rest-of-nation numbers, and "TotalCommoditySupply" to work with as the starting point. Anything left over after foreigna dn domestic exports would be used in-county. Sound right? What is the difference between the "TotalCommoditySupply" and the "TotalLocalSupply"?? Thanks for looking into the data discrepancies. I appreciate it.
  • Avatar
    ScottL
    Did you work with Mike Retzlaff to get that? I am not aware of us creating any non-traditional datasets nor any access queries. The way it works is we have total commodity supply. Foreign exports are removed to make net commodity supply and then we have local demand for net commodity supply, what is left over is domestic exports.
  • Avatar
    rochette
    We were working with Greg Alward for the 2007 data. All we received was a table of data that included all commodities and all counties in the US, a table that showed what the commodities were, and a table that showed what the county codes were. I created all the queries myself to get the numbers out of it that we needed, aggregated it myself and everything.
  • Avatar
    rochette
    [quote="Scott Lindall" post=8186]The way it works is we have total commodity supply. Foreign exports are removed to make net commodity supply and then we have local demand for net commodity supply, what is left over is domestic exports.[/quote] That doesn't seem to work... For El Paso County, commodity 3002, I have TotalCommoditySupply 0.66237200197164 ForeignExports 0.20836785435 NetCommoditySupply 0.4540041476148 LocalDemandforNetCommoditySupply (don't see this - what is the column name?) DomesticExports 0.45380180957 Are the columns you're referring to named something different than the ones listed? Should what you call "Local Demand for Net Commodity Supply" be "LocalSupplyToDomesticDemands" (0.000203966657864)? Actually, "LocalSupplyToDomesticDemands" and "DomesticExports" add together to "NetCommoditySupply". Thanks
  • Avatar
    ScottL
    DomesticExports = [NetCommoditySupply]-([TotalGrossCommodityDemand]-[TotalImports])
  • Avatar
    rochette
    [quote="Scott Lindall" post=8178]The main difference is in the construction flows. I will have to show this to Doug. He will be back Tuesday. I know there was an issue with this in a couple of states for 2009, but I thought it was fixed awhile ago.[/quote] Any updates on construction flow discrepancies? Thanks
  • Avatar
    ScottL
    Doug is looking at it.
  • Avatar
    ScottL
    Are you with the Forest Service?
  • Avatar
    DougO
    As far as the decrease in cross trade between 2007 and 2009 - it is related to a change in procedures to force services to be used locally. There were cases where the within county impedences were greater then impedence to nearby counties causing virtually all services to be bought in the neighboring county. That was not the case with the contiguous counties of El Paso, Fremont and Pueblo, but they were close enough to each other (in terms of intra and inter county movements) that the tightning of the services cause the difference you see.
  • Avatar
    eyosiyas
    What is the difference between the figures for domestic and foreign trade we get from ''IxC inpout output accounts'' and the ones we calculate from net commodity supply and total gross commodity demand? Both ways, do we get the same figures for domestic and foreign trade or are they different in meaning?
  • Avatar
    DougO
    Foreign trade is imports/exports from/to the rest of the world outside of the US. Domestic trade is imports/exports from/to the rest of the US.
  • Avatar
    eyosiyas
    How can we calculate these numbers for two regions connected through MRIO?
  • Avatar
    DougO
    Foreign imports/exports would not happen with linked regions. Domestic imports are calculated as discussed in this thread: http://implan.com/V4/index.php?option=com_kunena&func=view&catid=83&id=5477&Itemid=35
  • Avatar
    rochette
    [attachment=148]compare_2007_2009_epc.xlsx[/attachment] [quote="Doug Olson" post=8199]As far as the decrease in cross trade between 2007 and 2009 - it is related to a change in procedures to force services to be used locally. There were cases where the within county impedences were greater then impedence to nearby counties causing virtually all services to be bought in the neighboring county. That was not the case with the contiguous counties of El Paso, Fremont and Pueblo, but they were close enough to each other (in terms of intra and inter county movements) that the tightning of the services cause the difference you see.[/quote] It looks like the impedences were far too much, however. See the attached spreadsheet with the numbers collected with the Implan 2007 and 2009 data for trade within and between El Paso and Pueblo Counties. It really makes no sense that construction trade flows would go from $205.89million in 2007 to $0.01million in 2009. Other industry changes also do not seem reasonable. How would total trade between the counties go from $2,525million in 2007 to $492million in 2009? This is a major difference. It seems like the 2009 data is far too narrowed, making it appear to be near useless. If these numbers are correct, a single construction project constituted more than 3/4 of the total construction trade flow from El Paso County to Pueblo. We need these numbers to be real and realistic. Total construction trade from EPC to Pueblo is not at all realistic, nor is a total trade of $492 million. How do we work this out?
  • Avatar
    DougO
    Actually it does make sense that the construction trade would be close to 0, as by definition, new construction employment occurs at the site of employment. There should really be no import or export of new structures (repair and maintenance are based in an office and can travel). The construction industry will still import goods and services as needed for new structures, but "output" occurs at the construction site. The 2009 data are the better estimate.
  • Avatar
    rochette
    [quote="Doug Olson" post=8225]Actually it does make sense that the construction trade would be close to 0, as by definition, new construction employment occurs at the site of employment. There should really be no import or export of new structures (repair and maintenance are based in an office and can travel). The construction industry will still import goods and services as needed for new structures, but "output" occurs at the construction site. The 2009 data are the better estimate.[/quote] Very interesting, but what is the source of construction employment? For instance, many subcontractor and sole proprietor trades live in Pueblo county or Teller county but come to El Paso county when they have jobs in El Paso county. So the contractor from a neighboring county is paid for work (generating a sale) and the money flows out of EL Paso county. Then the contractor pays their employees who could be from any of the three counties (commuting pattern as opposed to trade flow). Saying the job site is in a county and generates employment only for that county does not account for normal trade flows. Another way of putting it, is that for construction, Implan is measuring where work is done. In that case, there would be no trade flow at all, since all work done in a county, is done in that county! Implan must have assumed all construction is non-basic employment for each county. Is there someone at MIG we could call to discuss methods so we understand and feel comfortable with the data we are working with or is the forum the more appropriate way to work through these questions? Thanks
  • Avatar
    DougO
    True, the employment and labor income (and all other value added) is recorded at site), and that employment may indeed take their money home. There is considerable discussion on the forum about handling in-commuting and particularly the possibility of imported labor for construction. V3 actually makes this fairly easy to handle. If you know that 85% of the construction employment is imported labor (15% local) then in the impact analysis event line, after you specify the "sales value" (- ie, the value of the construction project the software fills in the employment) employee compensation and proprietor income associated with that value of construction. You then change the employee compensation to 15% of the initial value. Only that 15% will be used to drive the local induced effect.
  • Avatar
    rochette
    That's fine if you are trying to estimate the impact of a single project where the contractor is located in the county in question, and hires employees from outside the county. What we are saying is that some construction firms in one county do work in other counties, and we would expect to see a trade flow as a result of that.
  • Avatar
    DougO
    Income for labor who reside in one county, yet work in another (out-commuters) shows up in the SAM as a payment by trade (column) to households (row) - ie, households receiving income from outside of the region. The value that shows up in the SAM will be for all industries and would be a net value. If a region is net in-commuting, the value shows up as a payment by employee compensation (column) to trade (row) - ie, importation of labor.
  • Avatar
    rochette
    It still does not make sense to me. If company A is located in Pueblo, and gets a contract to do a construction job in El Paso County, it is a trade flow from Pueblo County to El Paso County in Construction. The workers live in Pueblo County, drive to the job site in El Paso County, and get their paycheck from the company in the same county they live in: Pueblo County. It is not a household payment. It is an industry payment, who then pay thier employees just like any other firm involved in trade. Thanks for all the help in understanding.
  • Avatar
    DougO
    The BLS handbook requires that employers report employment by establishment. If the employer has multiple worksites, he is expected to file form MWR which divides up the employment to those various worksites for a reporting period. Each construction project is considered a worksite and the employment is to be divided up based on the work site. ( http://www.bls.gov/opub/hom/homch5_b.htm ) Employment, and associated income and value of production is workplace based - and in the case of new construction it is the site of construction. New construction is neither imported or exported, it is produced on site. Intermediate goods and services (concrete, sand, gravel, power) tend to be purchased from local resources - although this may be skewed towards the home contractor's administrative office contacts that they are comfortable with, but we have no data on that. Materiels used in production may be imported, or not, just like any other industry production function. Our underlying data sources do not indicate how much new construction use non-local primary contractors, nor does it indicate how many of the sub-contractors are based outside of the region as our data sources treat all construction employment as local. {Note, that use of construction subcontractors "assigns" that portion of the overall contract to that subcontractor. The subcontractor's employment and income is part of the direct effect (as opposed to an indirect effect). I-O definitions do not allow purchase of new construction as part of any industry's intermediate input as new construction is defined to be an investment. Production functions may only contain operational activities and do not include investment.}

Please sign in to leave a comment.