Jobs Report -- Don't Act Surprised
So, everyone is touting how wonderful the new jobs report is. It's the same as the previous 9 reports. If you go to the Bureau of Labor Statistics and look at the historical data (including revisions), and select:
1) Private Sector Data -- don't include government
2) Either Seasonally Adjusted or Not -- doesn't matter, same story either way
3) 12 month % gain
you will find an increase of 1.7+/-.1 % for the past 10 months straight.
I believe the variation is due to government layoffs, poor seasonal adjustment, and whether or not you count jobs in one month or the next. They average out to give a simple picture of moderate recovery.
The big test of how poor our seasonal adjustment procedure is will be one month from now. January has a huge drop in seasonal employees (not a surprise), so if we overcount or undercount them, they will look like January gains or losses. If the numbers are high, that might be because we overcounted seasonal jobs and vice-versa.
To convince me that something has changed, we're going to have to leave this 1.7+/-.1 % band.(*)
(*) Ideally, I'd use a geometric averaging process, whereby we go back several years with a geometrically declining weight each year. This avoids problems like a sudden seeming drop in numbers one year *after* the census because we're comparing to the census hiring. However, in the absence of large changes (like the census), one year % change is good enough.
1) Private Sector Data -- don't include government
2) Either Seasonally Adjusted or Not -- doesn't matter, same story either way
3) 12 month % gain
you will find an increase of 1.7+/-.1 % for the past 10 months straight.
I believe the variation is due to government layoffs, poor seasonal adjustment, and whether or not you count jobs in one month or the next. They average out to give a simple picture of moderate recovery.
The big test of how poor our seasonal adjustment procedure is will be one month from now. January has a huge drop in seasonal employees (not a surprise), so if we overcount or undercount them, they will look like January gains or losses. If the numbers are high, that might be because we overcounted seasonal jobs and vice-versa.
To convince me that something has changed, we're going to have to leave this 1.7+/-.1 % band.(*)
(*) Ideally, I'd use a geometric averaging process, whereby we go back several years with a geometrically declining weight each year. This avoids problems like a sudden seeming drop in numbers one year *after* the census because we're comparing to the census hiring. However, in the absence of large changes (like the census), one year % change is good enough.