Differences in Output per Worker FRB vs IMPLAN
2008 CONSTR sector Q-per-Worker is approx $151k. The Federal ResrvBank of San Fran (Center for Income & prodcutivity statistics) shows $81k for 2008 & cites BEA SCB for the source of data. I am inquiring since someone published impact results using a ReDYN model (back in 2008) and the inferred Q-per-Worker seemed low in comparison to IMPLAN. Then I saw the Federal Reserve statistics which is also closer to the ReDYN value. What is behind this difference? Thank you.
here's the link
http://www.frbsf.org/csip/data/charts/chart36a.cfm?init=1&selectedsector=Construction&state1=CA&selectedstate1=CA&state2=&selectedstate2=&state3=&selectedstate3=&selectedyear=allyear&datachoice=level&selectedregion=&selectedUS=US
-
Hi Lisa, We haven't found the reason behind the discrepancy but are looking into it. By the way, if you have any additional information about Fed Reserve Bank of SF's figures, that would be very helpful. Their data files only contain the productivity (i.e., output-per-worker); the output and employment figures used to calculate the productivity are not there and their link to the source is broken. This makes it difficult to determine what might be causing the discrepancy. At the very least, we will outline the methods/data we use to generate our figures. Thank you for your patience. -
Jenny - thanks for looking into it. As I mentioned, the FRB-SanFran web-site cites the BEA-Survey of Current Business for the data but I did not click the link (was it linked?) However, I just scoured the BEA web-site and under Annual Industry Accounts (updated 4/26/2011) the Gross Output Data by Industry (& employment) lead to amounts comparable to IMPLAN. -
Our construction output numbers: 1. State and Local public construction by IMPLAN sector comes from Census "Value of State and Local Construction" put in place for current year and is constrained to the BEA NIPA value for all state and local structure investment. 2. Federal parallels state and local 3. Private new construction comes from BEA NIPA table 5.3.5 Private Frixed Investment by Type. The output for each IMPLAN new construction sector is the sum of Public and Private investment. The output for the repair and maintenance construction sectors are a ratio of r&m to new construction from the BEA output by benchmark sector data released each year (usually lagged one year to IMPLAN data. All construction Wage and salary employment data comes from BLS employment security data for the current IMPLAN year. Proprietor data comes from REIS (lagged one year to IMPLAN data year). It is distributed to the IMPLAN construction sectors based on compensation to output ratios from the 2002 benchmark and the compensation per worker. Our thinking of the differences compared to FRB data may be that their output per worker is on a per project basis. If the average project lasts 6 months, then the average annual output per worker (IMPLAN basis) would be double the reported value. -
Rereading my response, I should note that I do not know if the FRB uses employment on a project basis. It is just a suggested possibility. Looking at the EB-5 applications rules, employment seems to be loosely worded: "2.Create at least 10 new full-time jobs either directly through the capital investment."[sic.. I assume it meant to say "...either directly or indirectly..."] I could hire 10 teenagers to weed my yard for 8 hours a day for every summer weekend. "Full-time equivalent" would better imply 2080 hours of labor a year for each full time job.
Please sign in to leave a comment.
Comments
5 comments