Reclaiming Common Sense

Seasonally Adjusted First-time Claims Should Have been a 19 Year High

The weekly unemployment claims number used to be a major headline every Thursday. The data recorded is the non-seasonally adjusted (NSA) data. The data that is reported is the seasonally adjusted (SA) data. The seasonal factors used to convert the NSA data to the SA data are different for the First-time Unemployment Claims (FTU) and the Continuing Claims (CC) data. The seasonal factors change from week to week, month to month, season to season, and year to year. The authors of the report have created FACTs (False Assertions Considered to be True) by comparing data from different years and different seasons with radically different seasonal factors. The largest FACT created was that we have had over 90 consecutive weeks with SA FTU claims under 300,000 people. Two weeks ago they pegged their FACT at 97 weeks. Last week they were silent. We did have a seasonally adjusted FTU number under 300,000 last week. Why no mention of the streak? Could it be that it didn't really start until January of 2016? Could it be because it ended Thanksgiving Weekend of 2016?

So what was recorded this week and what was reported?

Was it the Best NSA First-time Claims Number or the Worst SA First-time Claims Number or Both? The NSA FTU Claims number dropped from the upwardly revised 350,307 people to 280,463 people. This is slightly lower than the 281,885 claims recorded during the week ending 1/24/2015. (The SA FTU number back then was reported as 267,000) The SA FTU data was reported at 259,000. Eight thousand claims does not seem to make a big difference.

What would the Market Response Be if the FTU Number was reported at 183,000? We had the lowest NSA FTU number for the third week of January during the past 18 years. The authors of the report have manipulated the seasonal factors used to convert the NSA data on a regular basis. This is why the FTU Streak is Fake News.The data for January 7, 2016 was revised downward to 412,311 from 412,230. This is still higher than the advance number. The data from last week was also revised higher. Will this week's data be revised higher next week?

What would the response have been if it was Reported as the Highest SA FTU Claims number during the past 18 years?  This is probably the best display of lying with statistics that has been published in this column. They have lied about the true seasonally adjusted Current Employment Statistic (CES) "Jobs" number for years. They have been lying with statistics regarding the SA FTU number for years. It has never been so clear how much manipulation has been done until you look at the numbers and see that the lowest NSA FTU number for the third week of January, since January 2000, should have been reported as the highest SA FTU since January 2000.

The Continuing Claims Number Continues to hover around 2.5 Million Claims. The data for this week was recorded at an advance value of 2.461,553. The advance data from last week was revised from an advance value of 2.501 million to a preliminary value of 2.518 million people. The data published two weeks ago was revised up for a second time from 2.463 million to 2.467 million people, We are at a eighteen year low for the second week of January. The CC data trails the FTU data by one week.

The weekly unemployment claims number is being recorded and reported at artificially low levels. They are artificially low because of the seasonal factors used to convert the data from NSA to SA. They are artificially low because the participation rate is regularly at or near a generational low. We are missing millions of participants. Participants are those who are either employed or unemployed. If they are not employed they are either retired or not officially unemployed. The workforce is graying. The non-participants are not retired - they just aren't employed. The unemployment claims number is artificially low because those people who are recently unemployed  may have been working two jobs. They aren't unemployed - they are less employed. There are no "less employed benefits."

Next week is President Obama's final Jobs Report. Others will tell you that it is President Trump's first jobs report. They are wrong. The data collection for this January jobs report ended on January 12th. January 20th was the inauguration. We have one more Obama Jobs Report and a "Four Presidents at 96 months Column" coming up over the next two weeks.

Unemployment claims numbers have been an important economic indicator. We may be in a jobs recession. The data may not catch up with reality until it is too late, in part because of seasonally adjusted data and in part because of the changing balance in the workforce. We have been shedding full-time jobs, most likely seasonal, since July. We shed some multiple job workers last month. January is a non-seasonally adjusted net Current Population Survey (CPS) job loss month and a net non-seasonally adjusted Current Employment Statistics (CES) private sector worker loss month. The January Jobs forecast article will address the magnitude of the changes.

It's the economy.