One thing always to remember with the BLS (and the ADP) numbers is that they aren’t “real.” I don’t mean they are the product of a conspiracy. But they are generated by means of three surveys (two by BLS and one by ADP) which suffer from all kinds of methodological problems. For example, if you worked 5 hours last week, you are counted as employed. If you desperately want a job but haven’t seriously looked in 12 months because you kept being turned down, you are not counted.
Moreover, the numbers aren’t “real” because they apply seasonal formulas smooth out the huge peak in holiday hiring at this time of year and the trough that comes in January and February. And there is the “birth-death” of businesses formula. None of this is to say there is a conspiracy. But anybody who says without qualification that the economy generated xxx,000 jobs in October will be deluding themselves or their audience. And there will be plenty of reporters doing exactly that. It just isn’t that exact. It’s part science, part art. And it gives us the trend, which since March has been on a plateau that isn’t good enough to put all the millions of people who are out of a job back to work before 2017, much less absorb all the new entries to the working-age population. Even the “October” part is not quite accurate: The BLS report actually measures Sept. 13-Oct. 12.
MeteorBlades commented on the blog post BLS May Delay October Jobs Report Due to Frankenstorm
There is no need to worry about the survey itself. Surveying is done the third week of the month (data closes the 12th of each month, meaning that we are actually measuring not October but mid-September to mid-October). The only thing the BLS folks need to be in the office for is the analysis of the survey.