Sector 429 in 2012 Data

Comments

3 comments

  • Avatar
    IMPLAN Support
    Hello, Some changes occurred when we released a second quarter update to the 2012 data. We incorporated improvements and a newer raw data source, as outlined in the attached public release notes for those data. https://implan.com/index.php?option=com_content&view=article&id=899:econ-jenny-update&catid=14 We also want to ensure that you have an updated Structural Matrix: http://implan.com/index.php?view=download&alias=271-implan-structural-matrices-2013&category_slug=2013-data-information&option=com_docman&layout=table&Itemid=1480 Please remove your current Structural Matrix first before replacing with the one attached to this post. Structural Matrix: C Drive > Users > Users Name > App Data > Local > MIG > IMPLAN > System Data. Place the Structural Matrix file in the system data folder. However an issue with the Employment in government Sectors did occur in the second release. This link describes the basis of the issue and a fix you can apply, http://implan.com/index.php?option=com_kunena&view=topic&catid=80&id=18893&Itemid=1679#19263 or if you prefer we will be happy to replace you 2012 R2 data with R1 data. Please contact us at 651.439.4421x2 to replace your matrix to confirm if it's correct in if you want to keep the R2 or to replace you data if you are interested in having R1 data instead.
    0
    Comment actions Permalink
  • Avatar
    speters
    Hi again. Thank you for your response. The 0.16 adjustment seems to work for Direct Sector 429 Employment, as suggested. However, after fixing that we noticed that our aggregate numbers fell considerably between the 2010 totals and the 2012 numbers. After investigating further, we have found some interesting anomalies in the data. Direct Employment seems fairly consistent from 2008-(adjusted) 2012. However, Output in 2010, Intermediate Expenditures in 2009 and 2010, were MUCH higher than for other years, and Employee Compensation varied considerably over the period we looked at. Attached are some numbers we got for the U.S. model, Sector 429. Are there known reasons that these numbers might have varied so much over 2008-2012? For example, we would not necessarily expect changes in classification of construction activity to account for these year-to-year differences, especially in light of the lack of noticeable trends. Were there other methodological changes during that period, or do you have evidence that the classification of spending is in flux? The recession may have caused some changes, but would employee compensation have more than doubled from 2010-2011 while Output declined by more than 50%? We think the result of these large changes in interim expenditures and output caused the major changes in the indirect and induced multipliers between the years. For example, for the United States model the indirect and induced employment multipliers were 5.79 and 8.16, respectively, in 2010. Those same multipliers were 1.3 and 5.4 in the 2012 data. In our own estimates of employment from Federal OCS oil and gas activities, we found a huge drop from 2010 to 2012, despite similar input levels. The fact that we allocate spending to a large group of supporting sectors would seem to rule out the influence of any one sector definition or reporting changes having a large impact, except for government spending. Any advice you could give us on why there were huge variations in data between the years would be very helpful. Thank you, Sarah Peters
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Support
    Thank you for your questions. In general, yes, methodological changes do account for some of the shifts you are seeing, as do changes in source data. In the time period from 2008 to 2012, we made some methodological changes in allocating EC and Output among federal enterprise sectors (including postal services, utilities, and other enterprises). There was an exaggerated growth rate in output for sector 429 from 2009 to 2010 due to these changes. Most of this growth was allocated to higher intermediate expenditures. Estimated output returned to a more expected level in 2011. The high growth rate of EC for that sector from 2010 to 2011 (with subsequent decline from 2011 to 2012) is largely due to a shift in compensation paid to federal postal workers (2010 to 2011) and to a time-series shift in the Bureau of Economic Analysis (BEA) National Income and Product Accounts (2011 to 2012). Regarding the 2010 to 2011 shift, our estimate of total compensation paid to all federal enterprise workers (including postal services, utilities, and other enterprises) remained constant, but the compensation flowing to postal workers specifically fell by approximately $6 billion. Most of the fall in postal worker compensation was allocated to sector 429 (other enterprises), as the control total remained largely constant. For the 2011 to 2012 shift, specifically, total employee compensation for all federal enterprises, in 2011, was estimated to be $61.3 billion. After that, the estimate was revised downward to approximately $57 billion for 2011 and $56.7 billion for 2012; this reflects a very small real shift. Since our estimate of federal postal and federal utility compensation between those periods was constant, the reduction was applied to EC in 429. In general, IMPLAN can use only the data available to it at the time, and since some of the source data providers (e.g., the BEA) revise historical estimates, treating IMPLAN as a time-series can result in large year-to-year shifts that do not reflect changes in the real economy. Accordingly, we do not recommend using IMPLAN as a time-series. However, we are working to create time-series data in the future. Please let us know if this addresses your concern of it you have any additional questions.
    0
    Comment actions Permalink

Please sign in to leave a comment.