Reclaiming Common Sense

We can't seasonally adjust our income,

why should they adjust the jobs data?

Yesterday was a difficult day on Twitter. People accused me of Lying with the Jobs data and lambasted me for 24 hours. Here's the thing: Don't Shoot the Messenger. Kings and Generals used to shoot their servant messengers if they did not like the news. We saw the Jobs Report Data for September  released Friday. It disappointed people and did not cause concern because the seasonally adjusted data was close to expectations.  It was so bad that I had to write a Sunday Column.

The column "September Jobs Report - Down is Up" detailed how we lost 1.1 million unadjusted full-time jobs, gained 1.3 million unadjusted Part-time jobs, lost unadjusted unemployed workers, grew the population, and saw participation drop. The problem is that all of that data comes from the Current Population Survey. The Current Employment Statistics data is used for the President's "Jobs Number." The jobs number of actually a "worker number" that double counts people working two jobs, and we have millions of people working two jobs. The data recorded is the non-seasonally adjusted (NSA) Private Sector data. The reported value is the seasonally adjusted (SA) data. The seasonal factors change from month to month, season to season, and year to year.

We Lost 475,000 unadjusted private sector workers last month.  This is not surprising. It happens every September after everyone has returned to school . It happens. Just as there is a surge in hiring for the Christmas Season, there is a surge during the pre-Summer months. The graph and data left of this paragraph come from

We lost a comparable percentage of people during September 2015. This was expected and projected in the September Jobs Forecast Column.'It was even expected that theseasonal factor would rise. The problem is that even though we lost 475,000 workers it was reported that we gained 167,000 workers.

 If we just used last year's seasonal factor we would have only added 120,000 workers. If we used the factor from the last Presidential election we would have added only 33,000. They had to use a seasonal factor never used during the month of September to gin up the jobs number.

We can't seasonally adjust our income, why should they adjust the jobs data?