Quick question.  

I understand that the tax models are generalizations and probably Copernican in nature; however, there appears to have been a overall and substantial drop in taxes per employee across the board between 2014 and 2016.

Was there a modeling or data source reason for this - or was there a massive tax cut that my accountant did not hear about?

Thanks much

Was this post helpful?
0 out of 0 found this helpful

Comments

1 comment

  • Official comment

    Hello,

    Changes in our data creation process do contribute to the difference in taxes from the 2012 to 2016 IMPLAN data set. I've included some information on these changes below, but first I want to note few things after reviewing the spreadsheets you provided:

    • Not all taxes are associated with employment, so the ratio of taxes to employee count will not necessarily remain constant over time.
    • Because Dividends are not taxes they should not be included in your calculation of taxes. 
    • As we compared the 2012 and 2016 taxes in your spreadsheet with the taxes in our 2012 and 2016 models in IMPLAN Pro, they did not match, so we'd like to get a little more information on what IMPLAN platform and data you are using. Are your using IMPLAN Pro or IMPLAN Online? Is this data from our Time Series data set? I notice the 2012 and 2016 model tax categories are the same, so this suggests to me this is Time Series data. With this information we can further investigate. 


    Looking at the 2013 release notes, you'll see that the second release of 2013 data reflects the most recent (2012) 5-year Census of Governments, which improves estimates of government spending and revenue.  2012 and 2016 straddle different sides of that line, which is probably the main cause of the differences.  There were also many other significant changes that happened to our data in 2013: https://implanhelp.zendesk.com/hc/en-us/articles/360000751014-2013-Data-Release-Notes

    Please let me know if you have any additional questions!

Please sign in to leave a comment.