Showing posts with label bls. Show all posts
Showing posts with label bls. Show all posts

Sunday, June 7, 2020

Calling BS on the Unemployment Rate

On Friday U.S. Bureau of Labor Statistics released their monthly headline unemployment rate based on their household survey and it came in at 13.3% down from 14.8% in April. The release was followed by moronic headlines like....

"Economists predicted 20% employment in May. How did they get it so wrong?"

"Economists have biggest miss ever in U.S. jobs-report shocker"

However, the reality is that these headlines should have been written about how wrong the SURVEY WAS!

The published rate is straight nonsense and can EASILY be proven so.....and not only is their published number nonsense, so is their disclosed error. While their disclosed misclassification is a step in the right direction, it's completely inadequate. Below is the disclosure at the bottom of their press release.


So essentially we know the 13.3% is BS, but I'm also saying the claim that unemployment is only 3% higher is also BS!

This is easily provable by just looking at the governments OWN numbers. We don't need a 'survey'. Obviously a subset of the total unemployed will receive unemployment insurance and this number can be found from the Department of Labor. On June 4th, they released their most recent advanced figure on claims. Finalized numbers on all people that received unemployment benefits are reported with a 2 week lag, so this report showed those claiming Unemployment Insurance benefits as of the week ending May 16th. This roughly lines up with the date of the household survey, which is done around the middle of every month. Below is a look at the number of unemployed based on the household survey vs the number of unemployed based only on those receiving unemployment benefits (This is primarily made up completely from those receiving regular state unemployment benefits and Pandemic Unemployment Assistance). Here is what we see since the start of the year. I used the 3rd week of every month as the date to represent the household survey and filled the in between dates using trend.

And here is another look at that data but translating that to an implied unemployment rate rather than total number of people

As you can see, obviously not every unemployed person is going to receive unemployment insurance and generally the total number of unemployed had been 3-4 million higher which translated to about a 2-3% higher unemployment rate than implied by solely looking at those getting benefits. And as you can see that trend stood even after the new Pandemic Unemployment Assistance went into effect in April. 

It is CLEAR that this most recent unemployment report based off the household survey is complete BS. Looking only at insured unemployed we can tell that total unemployment went UP dramatically from April's report to May's report. Unemployment only started dropping slightly starting May 9th. 

Even if we were to conservatively and stupidly assume that every unemployed person receives benefits then the unemployment rate stood at 18.9% around the time of the household survey....not 13.3%. And if we added the regular 3-4 million uninsured unemployed that existed in every prior report this year we would have an unemployment rate more like 21-22%.

So let's just call this household survey what it is.....Bullshit.