This week, the first week of the month, we saw the release of the monthly employment situation report. This means that this column released its Jobs Report Prediction column. The first week of the month also means that this column releases it Top Ten articles of the month column for the prior month. The top ten columns focused on jobs and housing plus two of the week in review columns. Thursday was the Unemployment Claims Report. You didn't hear anything about the report because it was unbelievable. Friday we saw the release of an even more unbelievable report - the Jobs Report. If it seemed to be too good to be true that is because it was too good to be true.

(August 1) "The Top Ten Articles of July - The Private Sector Job Streak Ended - Housing Still Sluggish" revisited the column that compared the Four most recent two term Presidents at the 89 month mark, Part 1 of the Jobs Report Column that revealed we lost Private Sector Jobs during May, The combined April and May New Homes Sales Column, the combined April and May new Construction numbers column, The GDP column for the Second Quarter GDP and the week in review columns for July 9th and July 23rd.




(August 2) This column has written numerous columns on how it is easier for the authors of various reports to massage the data than change a long-standing message. This is very true for the monthly jobs report. The seasonal factors that are used to convert the non-seasonally adjusted (NSA) Current Population Survey (CPS)  data to the seasonally adjusted (SA) CPS data are not reported for the full-time jobs number, the part-time jobs number, the number of unemployed. The seasonal factors are also not released for how to convert the NSA Current Employment Statistics to the SA CES statistics. Growth rates are calculated for each category of data. Seasonal factors for prior years are calculated. The combinations and permutations of growth rate and seasonal factors are calculated.  The net result is that it was projected that we should expect a weak July Jobs Report. It was projected that NSA unemployment should  rise, full-time jobs should increase, part-time jobs should decrease, and that the participation rate was a crap-shoot. It was also projected that the Jobs streak number should have been around 150,000 to 160,000 and that it could be in the sub 100,000 range.

(August 4) Thursday we saw the release of the weekly unemployment claims report. We saw the lowest non-seasonally adjusted first-time unemployment claims for the month of July since 1967.  Yes Nineteen Sixty-Seven. Just read the column. The first-time unemployment claims streak is too good to be true. The first-time claims level is too good to be true.

(August 5) While the rest of the world was breathing a sigh of relief with regard to a "strong" July Jobs Report I was assessing how long it was going to take to sift through the data and find the real story. The seasonal factors used to convert the recorded NSA CES data to the reported SA CES skewed the data higher than it should have been. NSA Unemployment did increase. NSA Participation Increased. The "Official" Private Sector Number was that last month's NSA CES data was revised down from 123.191 M to 123.177 M and that the May Number was revised down from 122.105 to 122.099. Somehow a downward revision of 6,000 NSA CES jobs caused the SA CES for May to increase by 5,000. The streak still ended. Don't believe me - look at the White House Tweet. That third column from the right is negative. The problem with the 217,000 number is that if we used last year's seasonal factor the Private Sector number would have been 155,000.

If the data seems too good to be true it probably is too good to be true. It is also almost certainly seasonally adjusted.




PS. How was this tweet released at 7:42 AM when the report was released at 8:30 AM?