Proofreading yesterday's post after posting it -- shame on me -- I had another thought.
One might argue that the Minsky-Keen date for the end of the post-WWII golden age, 1966, stands out as a high point in the pattern of the "Average Annual Hours Worked" graph:
 |
| Graph #1 |
That
being the case, one might wonder if a comparable point in the pattern
-- the high point in the year 2000 -- marks the end of a somewhat less
"golden" age and the beginning of the age that sucks. You hear people
talk sometimes of "the end of the American century". This could be that
-- and, oddly, precisely on schedule.
Interesting, I think, because it puts the start of the 2007-2010 disruption (and related problems) at the year 2000 and explains, for example, the inexplicable 2.0% average Real GDP growth of the 2001-2019 period.
//
Why the decline in hours worked?
To
be sure, average annual hours worked depends at least in part on the
preferences of labor. But the guy whose preference is to work less that
the boss wants is soon not working at all.
The decline in average annual hours worked was no doubt in part a result of labor union efforts. But after the Air Traffic Controllers got fired, union efforts have had little to do with it:
A big part of the decline in average hours worked can be attributed to employers' preferences, which depend on economic conditions. When economic conditions deteriorate, average hours fall.
6 comments:
It looks to me that the labor force participation rate(LFPR) accounts for the changes in hours worked.
If you assume 1150 hours as the constant average hours worked by the entire population used to compute LFPR for the last 70 years you get close to the same graph.
The period from 1966-1980 was when many Baby boomers and women entered the work force. More workers working meant less average hours per worker. Shifting to more part-time and temporary workers was a strategy for cutting labor costs for many employers
https://fred.stlouisfed.org/graph/?g=zJ6u
Other things unchanged, more people working means less hours per worker. It's great you can put that on a graph right away and show it.
I looked at "Hours worked by full-time and part-time employees"
https://fred.stlouisfed.org/series/B4701C0A222NBEA
and at "Business Sector: Hours of All Persons"
https://fred.stlouisfed.org/series/HOABS
Both show a pretty straight path from the 1961 recession to the 2001 recession. Supports the view that changes in "average hours worked" resulted from changes in the LFPR.
If you divide "Hours worked by full-time and part-time employees" by "Civilian Labor Force Level" I would think that the result would be the same as your original graph of
average hours worked, but it doesn't look the same. It does still show a big decline in average hours from '69-'82
https://fred.stlouisfed.org/graph/?g=zRir
You get a high around 1951, a high around 1966, a high around 2000, and a high at the end, same as the "average hours" graph.
You get a low around 1958-1960, a low around 1982, and a low around 2009, same as "average hours".
And they both run flat from 1983 to 1993, give or take.
Also, the downward slope is very similar to the slope of average hours.
https://fred.stlouisfed.org/graph/?g=zRPs
Can't say about the slope of the green line as it is on the other vertical axis.
I see you multiply by 1000 to convert millions to thousands. I did the same and my red line ran an inch or so below the blue. I increased the constant to 1150 to move the red line up -- probably made the highs higher and the lows lower at the same time.
Oops.
... the same 1150 that you used above.
Curiouser and curiouser.
My 1150 came from multiplying labor force participation rate(LFPR) times average hours. This hovers between around 1100 and 1200 and stays close to 1150 for 66-82. That should be the average hours per capita (assuming the same workers were used for both data series)
This shows that a constant number of per capita work hours being divvied up between more workers in the labor force was why the average hours declined.
Your most recent graph shows that different accounting methods were being used for hours and or workers in the different graph lines that should both show average hours worked.
Post a Comment