Employment and EC numbers

This question arises often when IMPLAN data are compared to CEW. BLS QCEW data do not provide full coverage: It only covers wage and salary employees covered by Unemployment Insurance and federal civilian jobs covered by UCFE (Unemployment Coverage for Federal Employees). This means that military and railroad workers (who have their own retirement systems), religious organizations, elected officials, and some private elementary school jobs are missing from BLS employment counts. The BLS QCEW data also does not include proprietors (i.e., self-employed persons). For these missing components, we turn to the BEA’s REA (Regional Economic Accounts) data. We then ultimately control all estimates to the BEA’s U.S. NIPA figures, as these datasets attempt to capture all employment. IMPLAN Labor Income = Employee Compensation + Proprietor Income. EC is loaded payroll (that is, it includes benefits, payroll taxes, etc.). Proprietor Income can be negative. In contrast, CEW data only capture wage and salary income (no taxes or benefits) and does not include proprietor income - so they really cannot be compared! This webpage contains links to number of articles describing our process for developing employment data: http://implan.com/index.php?option=com_content&view=article&id=821&Itemid=1555 Please let us know if you have any further questions. Thanks

Hello, I do know the differences and your sources, I just quoted the QCEW numbers I had at hand as good suggestion that there’s an issue. I would appreciate if you would reference the particular industry I’m referring to and the study area data I attached to my email. Since you’re giving me boilerplate about non-inclusion of military and private elementary school jobs, etc., I think you have not. If QCEW shows an average annual average salary of $162,007 and IMPLAN shows employee compensation (which I know is salary plus benefits and taxes) per employee to be less than $24,000, it seems like there may be some kind of problem. Is either $24,000 average employee compensation or $34,000 labor income per worker really plausible? Payroll employment in the sector is about 4,200. You can then add nonemployers to this and that’s another 2,100 (per the Census Bureau). Is there really anything that would account for the Austin metro having employment of 33,295 in sector 356? For the larger finance and insurance industry (NAICS 52), QCEW gives 67,061 jobs and the BEA gives a little higher number, 69,768. I don’t see 3-digit employment in the BEA’s regional data. Even if all of the 2,707 delta happened in NAICS 523, that still wouldn’t get anywhere near 33,295. How many metropolitan areas would have “Securities, commodity contracts, investments, and related activities” as their 6 largest IMPLAN sector by employment? 3% of our total jobs? We had an issue like this in 2009. The issue was identified in your office as an error isolated to the government employment numbers, I was quickly given a work-around, and subsequently received new files for my counties and a new structural matrix file. Please ask the Econ-Team to give this a closer look and let me know what is found. I have attached the Excel file that summarizes what I’ve found. Thanks very much!