The problem I have with employment data is that it cannot be validated with any precision in real time.
You would think in 2016 with internet connectivity there should be no issue assembling accurate data in real time. Yet, the BLS uses two forms of ESTIMATING employment that are straight out of the dark ages:
- the household survey which surveys 50,000 to about 60,000 eligible households EACH month. The BLS says that for monthly change in the household survey employment to be significant, it must be plus or minus 500,000.
- the establishment survey sample covers over one-third of total employment universe. An over-the-month employment change of about plus or minus 100,000 is statistically significant
Whoa. That is a huge accuracy window for a piece of government data that the markets consider to be a very important document. Is the meaning of these sentences designed to be purposely ambiguous?
