For additional guidance on this topic, visit the IMPLAN Support Community at https://support.implan.com or contact IMPLAN support directly.

IMPLAN Sectoring

dahalram

February 13, 2015 19:04

Hello, I know IMPLAN sectoring is linked with BEA IO data and gets modified once the benchmark is modified. But how can we say both implan models are similar except for the dataset (say model built from 2002 data and 2010 data) during comparison . During the modification, some IMPLAN sectors get dis-aggregated and some get merged. For instance, sector 125 (2002 data) dis-aggregated into 105 and 106 (2010). If we run the impact analysis for these two years, how can we argue that the results are comparable? Aren't the results sensitive to different model parameters?

Was this post helpful?

0 out of 0 found this helpful

Comments

3 comments

IMPLAN Support
February 16, 2015 13:53

Hi Ram, Your thinking is exactly correct, impacts in Sectors whose definitions have changed indeed won't be perfectly comparable, and is a large reason why we caution time series modeling. I have included this article which briefly discusses modeling in a time series: https://implan.com/index.php?option=com_content&view=article&id=650:650&catid=253:KB33. We recommend using the ES202/CEW data for time series analysis: https://implan.com/index.php?option=com_virtuemart&view=category&virtuemart_category_id=10&Itemid=1647

0
dahalram
February 18, 2015 21:16

Thank you for the reply. Does the bridge table provided by IMPLAN to relate 509 sectors to 440 sectors solve the problem of direct comparison? Are the IMPLAN IO model same for 2002 (say) and 2010 except for the data set? What I mean to say is if we run the combined impact analysis for sector 105 and 106 for 2010 and compare this result with the impact of sector 125 for 2002, am I doing the correct job? Thank you.

0
IMPLAN Support
February 23, 2015 17:24

Hello, IMPLAN is unfortunately not designed to be used as a time-series. This is because with we update some background calculations and data sets as we find better methodologies or sources. One large change is the Trade Data. 2007 and forward IMPLAN Data years Trade is estimated with the Trade Flows. You can solve this issue by building all your models through the Econometric RPC method in User Preferences>Social Accounts tab.

0

Please sign in to leave a comment.