I've been watching
unemployment data
since the 1970s.
I prefer the Household
Survey, which does not
need revisions,
while the headline
Payroll Survey
tends to have large
negative revisions
just before and just
after a recession starts,
making it misleading
just when you want it
to be accurate !
The Household Survey
collects far less data,
so is about 50%
more volatile than the
Payroll Survey,
which means
averaging two or three
months of Household
Survey data is mandatory.
In a June 11 article,
I told you the
Household Survey
I told you the
Household Survey
for May 2019 reflected
slightly fewer jobs
than in Novemebr 2018,
That was very bad news.
But ... some optimists,
who never expect
a recession, told me
the May 2019 Payroll
Survey jobs gains
may have been small,
but that survey
completely contradicts
the Household Survey.
MAY 2019
EMPLOYMENT
"PAYROLL SURVEY"
The +75,000 new jobs estimate
for May 2019, was obviously
bad news.
I instinctively look at
prior months revisions,
in addition to the latest
monthly jobs data.
I doubt if many other
people bother with that !
Add the March 2019
and April 2019
negative revisions,
to the initial May 2019
estimate, and the total is
NO ADDITIONAL JOBS,
beyond what was reported
one month earlier !
Negative data revisions
are an important symptom
of a weakening economy.
Few people seemed to realize
revisions for the prior two months
subtracted a total of 75,000 jobs:
March 2019 was revised down
from +189,000 to +153,000,
a loss of 36,000 jobs.
April 2019 was revised down
from +263,000 to +224,000,
a loss of 39,000 jobs.
The March And April job loss was
(-36,000) + (-39,000) = (-75,000)
Note:
The average jobs growth
per month in 2018
was 223,000.
The monthly jobs growth average,
for the three months ending May 2019,
declined to 151,000, which was
the smallest 3-month gain in 20 months.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.