(Economics 100b; Spring 1996)
Associate Professor of Economics, 601 Evans
University of California
Berkeley, CA 94720
(510) 643-4027 phone (510) 642-6615 fax
January 26, 1996
Scattered Reflections on Pedagogy
The Idea of "Unemployment": Cyclical, Frictional, and Structural
Some Unemployment Magnitudes
Next Time: Production and Distribution When the Economy Is Near
The Idea of "Unemployment": Cyclical, Frictional, and
On May 23, 1995, I resigned from the Treasury Department. On June
1, 1995, I went onto the U.C. Berkeley payroll. For those eight days
in between, I was "unemployed". Similarly, my brother's clerkship for
the Hon. Ralph W. Winter, Judge on the U.S. Court of Appeals for the
Second Circuit (our family judge, in fact) ended on July 1, 1995. Yet
he did not accept another job until he went to work for the New York
law firm of Howard, Darby at the beginning of September.
It is an abuse of language to call me "unemployed" at the end of May,
even though I really was unemployed according to the Bureau of
Labor Statistics' definition. My brother was more genuinely
"unemployed"--although he had at least three law firms begging him to
please come and work for them during July and August.
My brother was unemployed because he was hoping for a better job
offer to come along--and it did. I was unemployed becasue it is a
real pain in the to move across country. Both of us were examples
of what you might call "frictional" unemployment--the fact that
we were unemployed did not mean that we were suffering particularly,
did not mean that the economy was malfunctioning in any sense, but
was instead simply part of the normal process of people finding and
changing jobs, just as the fact that every night a large number of
full cargo containers are stacked up at the port of Oakland does not
mean that our system for distributing goods to consumers has broken
down: "inventories" of goods and of labor that are not "in
production" at any particular point in time are a necessary grease to
keep the economy operating smoothly.
However, there are other kinds of unemployment than
I once had dinner with a woman who was leaving for England on an
assignment the following day: she had been hired--this was the
mid-1980s--to retrain ex-coal miners in the north of Great Britain so
that they could get jobs as clerks and cashiers in local shopping
malls. She thought that this was going to be an incredibly difficult
job--that the cultural heritage of the ex-coal miners whom she was
going to be retraining was close to being completely inconsistent
with the skills and attitudes required to be a good clerk or cashier
in a shopping mall.
- Cashiers need to be hyper-polite: say sir or ma'am; start from
the belief that the customer is always right; take abuse from
cranky customers without getting mad; sympathize.
- Cashiers need to follow directions: managers have little
patience with people who won't do things their way, quickly,
- Cashiers need to have good basic math sense: to figure
out--quickly--when the amount of change that the machine is saying
you should give this customer is clearly wrong; to be always
running sums and differences in their heads to make sure that they
have pressed the right buttons.
Now spending twenty years working underground in the
highly-skilled, -specialized, and -dangerous occupation of coal
mining is not going to predispose you to have the orientation toward
people--and the willingness to accept the requisite social role--to
be a good cashier.
But better machinery and more efficient mines elsewhere have closed
the mines in Britain's north; moving to another part of the country
is hard to do--especially because the structure of British local
government and housing subsidies means that your housing costs
quadruple if you move. And if you don't have the possibility of
adding value to a business by working in a mall--if the hassle to the
manager and your wage are, together, a bigger cost than is the value
of the work you do--it is hard to imagine what you are supposed to
You are "structurally" unemployed.
Over and above "frictional" and "structural" unemployment, there is
"cyclical" unemployment. You can't get a job in your usual line of
work. You are pretty sure that your unemployment is the result of
slack demand in your particular industry in that year. Switching to
another industry would be pointless: by the time you got retrained
and found a job, your old employer (or someone similar) would
probably be trying to hire you again.
Social balance sheet of types of unemployment:
- Frictional not a social loss--"too little" frictional
unemployment would be a social loss. People should
spend some time looking for jobs. The value of the "match"
- Structural a social loss, but not fixable through
- Cyclical a social loss
"Frictional," "structural," and "cyclical" as ideal types. They blend
into each other:
- Labor force upgrading in a high-pressure economy
- Labor force downgrading in a low-pressure economy
Some unemployment magnitudes:
- Highest (annual) unemployment since 1945--10% or so in
- Lowest (annual) unemployment since the end of World War
II--2.9% during Korean War
- Lowest (peacetime; annual) unemployment since the end of World
War II--3.5% at the end of 1969
- Average unemployment...--approximately 5-6%
- Unemployment by race, sex, and age. Black
unemployment roughly double white unemployment; male unemployment
used to be lower than female unemployment--but not any more.
Teenage unemployment quite high--two to three times "married
- Roughly half the unemployed have been unemployed for five
weeks or less; roughly 1/7 of the unemployed have been unemployed
for six months or more. Things are very different in
Europe--typically, in Europe 1/3 of the unemployed will have been
without jobs for more than a year...
- Unemployment by education: college-educated have
unemployment rates two or more percentage points below the rates
of people with just a high school diploma
Every month the U.S. Bureau of Labor Statistics picks a particular
week, and sends out its surveyors. They wander around the country,
and ask 60,000 households about their employment status.
Based on how you answer the questions--if you are one of those
surveyed--the BLS sorts you into one of three slots:
- Not in the labor force
A person is employed if he or she spent most of the previous
week working at a job--as opposed to keeping house, going to school,
doing something else, et cetera.
A person is unemployed if they wanted to work during that
previous week, but did not--because they were temporarily laid off,
because they have been hired but their new job has not yet started,
or because they were looking for work but did not find any.
Note that a person who would like to hold a job, but did not look
last week because he or she thought that it would be of no use, is
not unemployed. Such "discouraged workers"
are classified by the BLS as "out of the labor force."
Today the U.S. labor force is--roughly--133 million people, growing
at about 1.5 milion people each year. The total population 16 years
and older is about 197 million. Roughly two out of three
"adults"--people 16 or over--are in the labor force.
Today the unemployment rate is about 5.5 percent. Spurious precision.
Probably better to say 5.5 percent + or - 0.3 percent.
- Estimates of state-level unemployment rates have even bigger
margins of statistical uncertainty--of sampling error. The
California state unemployment rate, for example--now somewhere
around 7.5 percent--is really 7.5 percent + or - 1.0 percent.
The Federal Reserve worries when the unemployment rate is 5.5
percent. It is scared that inflation is about to accelerate--that
businesses will think that demand is so strong that they have to hire
more workers, and can increase prices to pay those workers higher
wages. But--so the Federal Reserve thinks, so most of us think--when
the unemployment rate is south of 5.5 percent, firms are apt to
mistake high aggregate demand for high relative demand for
their products--and think erroneously that they can rise prices.
On the other hand, the Federal Reserve has no desire at all to
keep unemployment above the sum of "frictional" and "structural"
Arthur Okun--Chairman of the Council of Economic Advisers under
President Johnson in the 1960s--noticed a striking statistical
- every one percentage point decrease in the unemployment
rate seemed to be associated with a 2-3 percent increase in
output relative to the economy's long-run potential growth path.
At first thought, this is kind of odd. A one percentage point
decrease in the unemployment rate seems like it ought to be
more-or-less a one percent increase in the number of workers. So we
would expect output to grow by at most one percent (maybe less if the
new workers cannot find sufficient capital goods to boost their
productivity up to the level of the previously-employed).
Instead we find--consistently--output growing by 2-3 percent.
- When the unemployment rate falls by one percentage
point, the labor force rises by about one
percent--discouraged workers re-enter the labor force.
- Hence a one percentage point fall in the unemployment
rate is associated with approximately a 1.5 percent rise in the
number of workers.
- A one percentage point fall in the unemployment rate is
associated with an 0.5 percent increase in the length of the
- And--perhaps--a one percentage point fall in the
unemployment rate is associated with an 0.5 percent increase in
average labor productivity. Production processes seem to
work more smoothly--less downtime, less waiting around, more
efficient use of fixed distributional resources, a whole bunch of
things that you can label increasing returns to scale.
Thus a one percentage point fall in unemployment is associated
with a 2 percent increase in labor input, and a 2-3 percent increase
This suggests that in a sense our measures of unemployment are wrong:
- People out of the labor force who would be in the labor
force if unemployment were lower are, in a very real sense, as
"unemployed" as anyone else.
- People who are parttime for economic reasons are partly
- Thus a more comprehensive index of unemployment and
underemployment would now, today, say that the nationwide
unemployment rate is not 5.5% but instead some 10% or so...
- How you evaluate this economic system depends on how seriously
you regard these various types of unemployment. You might accept a
5.5% sum of "frictional" and "structural" unemployment as not too
bad. But 10% somehow seems much worse...
Next Time: Production and Distribution When the Economy Is Near