How many people are really unemployed ?

Note: This is a longer than usual article, more like the articles featured in my former newsletter, ECONOMIC LOGIC. If you don't want to read it all, just read the INTRODUCTION and SUMMARY.

INTRODUCTION:
I am very confident 8.4 percent August 2020 unemployment is not even close to fairly representing the damage done to the US by COVID and government (over) reactions to COVID. The COVID pandemic is still in progress, so there are no experts yet.  Hopefully, in 2021 the pandemic will be history. And some people will have figured out the total economic and health damage done. I know that in the internet age, jumping to fast conclusions is popular, presented with scary headlines, based on incomplete, unverified data. Not here.

The real GDP decline of 10% in the first half of 2020 was huge.  No one will ever know how the U.S. would have done with no lockdowns, no social distancing, no HCQ medication, and no masks. I predict there will be lots of wild guess analyses, and politicians will claim their decisive actions saved a huge number of lives.

I have searched for months to find a decent government or private analysis of how many paid work hours were lost, relative to the early 2020 low unemployment level. No data analysis worth presenting here was found.

An alternative was to collect anecdotes and add them up. In medicine, that's called a field study. The fastest way for the CDC to find out what medications were working best for COVID patients, would have been by asking thousands of doctors for anecdotes. They failed miserably in that regard. One anecdote means little. Lots of anecdotes should add up to a good conclusion, long before results from a double blind test of any medicine would have been known.

So I used my imagination on this subject. The wife and I analyzed how everyone we know has had their paid work hours and income affected by COVID. Of course no names are used. We had plenty of time, being retired for 19 years, and 15 years, respectively!  I imagined these people being phone interviewed by the BLS, and answering the questions honestly. I also added more anecdotes from TV news interviews. The more anecdotes, the better the conclusion.  It was obvious the number of paid work hours lost were far higher than implied by the official unemployment rate.

In the DETAILS section I give 12 examples of people who are unemployed, or partially unemployed, but would not be counted as unemployed by BLS. Most examples refer to people I know, with no names mentioned. Imagining each person was called by the BLS for an employment survey, it was obvious that honest answers would have resulted in ALL of them not being called "unemployed".

Partially unemployed is a concept not included in the headline U3 unemployment number. That leaves out a lot of lost work hours, as if they do not matter. I say two workers whose paid work hours were cut by 50%, because of slow business during COVID, are equal to one unemployed worker. The BLS says they are both employed.


SUMMARY:
The Bureau of Labor Statistics (BLS) uses phone interviews to estimate the percentage of American workers with zero hours of work, who are actively looking for a job. If you worked one hour a week, you are employed. If you did not look for a job, you are not part of the labor force, as if you had retired. The traditional BLS methodology results in an extremely misleading view of the economy in the COVID age. The huge 10% decline of real GDP in the first half of 2020 is a better indicator.

29 million Americans are receiving unemployment compensation. Those receiving traditional state benefits have no work. Those receiving the new federal benefits may have no work, or sharply reduced hours. Those with reduced hours see themselves as partially unemployed. BLS sees them as employed,

I prefer the 29 million number as a better estimate of how many paid work hours were lost due to COVID infections, commerce reduced by customers fearing COVID,  government forced business shutdowns and voluntary resignations from fear of COVID, or to take care of children who can't attend school, or to take care of elderly parents who were living in a dangerous nursing home.

Starting in the spring, the BLS decided not to change their rules on how to determine the headline U3 unemployment rate in the COVID era. At the time the partial lockdowns were to "flatten the curve", and were expected to last one month, or maybe six weeks. Their decision made sense at the time. But the forced economic disruption lasted much longer than one month. And the traditional BLS methodology for estimating the unemployment rate makes the economy seem much better than it really is.

What I really want to know is how many paid work hours have been lost since the low unemployment rate in early 2020. I have not found any official analysis, or detailed analysis, that came up with a better "unemployment" number. The 8.4 percent Bureau of Labor Statistics (BLS) number for August 2020 gets 100% of the attention.

With 8.4 percent, Trump lovers are thrilled. Not that 8.4 percent unemployment is ever good news. But 8.4 percent is much better than the 14.7 percent in April 2020, and the declining trend makes a good narrative for Trump fans. The total number of Americans receiving unemployment compensation is also declining, but far less. The peak was 32 million in May and we are now at 29 million. It seems like Democrats don't want to challenge the BLS bureaucrats on their unemployment number, perhaps because they love "deep state" bureaucrats, most of whom are Democrats.

BLS was trying to be consistent with past years, which usually makes sense, but doing that makes the economy seem better than it really is. Which makes mandatory lockdowns seem not as bad as they really are.

I have been following BLS employment data since the 1970s, and only complained once in my former newsletter, ECONOMIC LOGIC. Their birth death (of businesses) model guesses the number of new jobs created at newly launched businesses. But using the model makes no sense during a recession -- that was my only complaint.  It was a minor complaint.

Government loans have kept the unemployment rate under some control -- companies borrowed money to keep paying employees they really did not need. I'm not discussing that charity now, although ... some very wealthy business owners got money, and many small businesses did not. The details of who got the money would hurt the Trump campaign, so we'll never know the details before the election.

If you will not be reading the DETAILS section, consider this one example:
-- A small business owner with much less business during COVID cuts all 10 employees paid hours by 50%, so their paychecks are cut in half. If the BLS called any of those employees, none of them would be binned as unemployed. If the business owner had laid off 5 of his 10 employees instead, which he didn't want to do, a BLS call to one of the five would have found an unemployed person.


DETAILS:
What we really want to know is how many paid work hours were lost due to COVID infections, fear of COVID, COVID partial lockdowns, and school closings. The official method appears to miss a large amount of unemployment, and partial unemployment.

The Bureau of Labor Statistics (BLS) calls people and asks how many hours they have worked, and if they have not worked, they are asked if they have actively been looking for work. Looking for work is something they would have to prove to keep getting state unemployment benefits. If they say "no", they are NOT officially unemployed -- they are booted out of the labor force count!  They are treated as if they are retired.

For August, BLS reported 13 million (8.4%) people unemployed using their traditional U3 report definition. A person who lost his job due to a COVID shutdown, but was not looking for a new job, because nothing is available, is officially not unemployed. A person who had his hours cut in half at his job is not unemployed. A person with two jobs, who lost one job, is not unemployed. All three people would think of themselves as unemployed, or partially unemployed, with no paycheck, or a much smaller paycheck.

This year there are new federal unemployment benefits aimed at independent contractors (aka gig workers, or 1099 workers) who were never eligible for (state) unemployment compensation. Add all the people receiving these benefits and there are 29 million Americans receiving state or federal unemployment benefits (18% "unemployment").

Those self employed gig workers can receive benefits if they are unemployed, or if they have had their hours cut significantly. Individual states may differ, but I've read that there are reduced benefits available for people who have lost income but are working up to three days a week, as long as they are earning less than $500 per week. (These federal benefits are quite small after the $600 a week federal bonus ended on July 31, 2020. They are even smaller if the recipient is still earning some money working part time.)  There may be very little work available, but if you work one hour during a week, the BLS says you are you are employed.

The EU is less honest about COVID related unemployment. The EU calls COVID layoffs temporary, and doesn't count them as unemployment, unless the business has closed permanently. They have the nerve to claim the 6.6% EU unemployment rate in January 2020 increased to only 7.2 % EU unemployment in July 2020. Very little effect from COVID !  Simple to do.  Don't count "temporary layoffs" as unemployment, Even though doing that stretches the definition of temporary.

Using federal government data, it seems impossible to know exactly how much COVID has reduced paid work hours. The 8.4% unemployment rate is just a subset of the total COVID lost jobs / lost hours effect. The 18% unemployment estimate, based on the unemployment compensation payments number, is probably more realistic. Because it includes some people forced to work fewer days per week. but still have some work. They are partially unemployed.

ANECDOTES:
Assuming the people in the following 12 examples were called by BLS, and answered their employment questions honestly. For most of these examples, I could use real names, but they are not necessary:
  NCAU  = Not Counted As Unemployed
  EFSFB = Eligible for Small Federal Benefits (small after July 31)

(1)
A small business owner
who has no work
= NCAU

(2)
A small business owner
who has a large
business decline
= NCAU

(3)
His laid off employees,
expecting business
to bounce back soon,
so are not looking
for another job
= NCAU

(4)
An older person
who voluntarily quits work,
or takes a leave of absence,
because the job requires
close contact with strangers
= NCAU

(5)
A chef, waiter
or a waitress,
who is out of work,
but expects to be
back post-COVID,
so he doesn't look
for another job,
since there are
none in the area
  = NCAU

(6)
A chef whose hours
are cut in half
= NCAU,
but maybe could
get EFSFB

(7)
A personal trainer
whose business
is way down
= NCAU,
but maybe could
get EFSFB

(8)
A person working
off the books
who loses his job,
or gets few hours
of work
= NCAU

(9)
A person getting
$4,000 a month
unemployment
checks, but not
looking for a job
= NCAU

(10)
A person who voluntarily
quits a job to take care
of her children, not being
allowed to attend school
= NCAU

(11)
A person who voluntarily
quits her job to take care
of an elderly parent,
who had been living in a
dangerous nursing home
= NCAU

(12)
A person who graduated
from school or college
this year, unable to find
their first real job
= NCAU

These examples, most based on people we know, show that you can lose a job, or have a large reduction of work hours, yet not be counted as unemployed. That's why I prefer the 18% unemployment rate estimate, based on 29 million (18% of labor force) Americans receiving unemployment compensation, over the official BLS 8.4%  unemployment rate. 

The 10% decline of GDP in the first half of 2020 suggests actual unemployment is significantly higher than the BLS reports. BLS has a COVID footnote to explain why their headline (U3) unemployment rate is not calculated differently in the COVID era. Their reason would have made sense if the partial lockdowns lasted one month. Just like we originally expected.

Meanwhile, partial lockdowns have lasted over five months so far, and there are still forced business and school closures. It seem that half the country wants the economy throttled until election day, to hurt President Trump's chances of reelection!  In addition, the mass media have frightened people so much that many voluntarily avoid airplanes, trains, busses and any business where six foot social distancing would be difficult or impossible.

COVID fear is what happened in Sweden, where there were few mandatory lockdowns. Only high schools and colleges were closed, not other schools, or businesses. But their economy declined in the first half of 2020 anyway, from voluntary social distancing. The fear of COVID, hyped by the mass media, will take a long time to disappear. Especially for older people like me. At this point it seems like the government (over) reactions to COVID were worse than the disease itself.



COVID  MEDICAL  FACTS 
 AND  ESTIMATES:
For COVID-19 infections, roughly 400 of 1,000 infected people seem to have no symptoms, or mild symptoms. Roughly 600 of 1,000 infected people have normal flu symptoms, or serious flu symptoms requiring hospitalization, and sometimes the ICU. About 1 of 1,000 infected people die.  I know that adds up to 1,001 of 1,000, so sue me!

Articles on the subject seem to focus on the 1 of 1,000, and ignore the suffering of the 600 of 1,000, where the overall suffering is much worse than ordinary seasonal influenza. The death rate seems small, but is actually large compared with a typical seasonal influenza.

Death certificates: Flu is not a direct cause of death listed on a death certificate.  Flu is not one of the over 100 causes of death on the official CDC list. The failure of a major organ is the cause of a non-accident, non-homicide death. If a flu infection is known to be present, that infection may be noted on the death certificate, or maybe not. "Flu deaths" are estimated by the CDC using computer models. Doctors think their estimates are too high. I imagine COVID flu deaths will be considered too high also.

Sometimes the only known health problem of the deceased was the flu. That doesn't mean there were no other health problems -- in fact, there probably were. Many people have high blood pressure, diabetes, heart disease or cancer present, but not yet discovered. A friend in his 50s had one symptom of severe heart disease -- his sudden death.  A friend's husband, around 70, had the first symptoms of pancreatic cancer just two months before he died.