Lecture Nineteen
Unemployment
(Economics 100b; Spring 1996)
Professor of Economics J. Bradford DeLong
601 Evans, University of California at Berkeley
Berkeley, CA 94720
(510) 643-4027 phone (510) 642-6615 fax
delong@econ.berkeley.edu
http://www.j-bradford-delong.net
March 4, 1996
Administration
Costs of Unemployment
Frictional Unemployment
Structural Unemployment
Cyclical Unemployment
Administration
Now that I have gotten through the first half of the course, I have
some additional thoughts about how to run the second half. My first
thought is that much of what I have to say about unemployment I have
already said--look back at lecture four. And that I should not repeat
myself (or if I do, I should repeat myself in closer proximity to
when I first said it). So I am going to compress discussion of
unemployment to--I hope--today alone, and start on the international
economy on Wednesday...
Costs of Unemployment
I want to stress, first, the costs of unemployment: making people
broke and pounding the pavement looking for jobs is not a fair way to
distribute the costs imposed by the business cycle. Unemployment is
perhaps the sharpest example of how the market system functions as an
economic discipline device: you fail to satisfy your employer, your
income vanishes...
Unemployment has costs in addition to the obvious losses of income
and security.
Some evidence that unemployment adversely affects productivity, that
is:
- Get fired once, it's hard to get excited about organizations
again.
- Without a job, you are likely to be depressed--and stay
depressed. To the extent that we live in a "partial gift-exchange"
world, productivity costs of unemployment can be very large.
Frictional Unemployment
Workers and firms spend time searching for each other. There are many
different kinds of jobs; many different types of workers. In a
complex economy, we cannot expect everyone to get matched to the best
possible job (or every job to get matched to the best possible
worker) instantaneously. Moreover, it may well take a considerable
time for people to learn about jobs and workers.
Moreover, over time skills change, the labor force changes, and the
needs of firms change.
Thus we should expect--and we should see it as a good thing--that
unemployed workers will not necessarily take the first immediate job
offer they see, and that firms should not necessarily hire the first
immediate applicant they see, but that instead time and energy are
spent in the process of search: trying to find a good match between
workers and firms.
This is frictional unemployment.
Recognizing that some frictional unemployment is a good thing is the
easy part. But how can we find out whether the U.S. economy has too
much, too little, or the right amount of frictional unemployment? We
can't, very well: it is unclear whether firms and workers spend too
much or too little time and effort in the business of searching for
each other.
Both firms' and workers' search decisions generate
externalities. When they find a "good match," they both are
likely to derive some benefit. Thus a worker who chooses to search
harder is likely to get a good match--but on the other hand is likely
to benefit the firm that he or she eventually chooses to work
for.
Thus a worker who keeps searching until the marginal gain to him or
her from searching an additional day is just smaller than the
marginal cost of searching an additional day may well be searching
too little from the standpoint of social welfare.
This is a reason to have an unemployment insurance system--in
addition to the "insurance" motive, there is a "job search" motive to
have such a system. This is also a reason to spend money on job
search assistance.
Does unemployment insurance lead people to search more? Yes.
Digression on Bush's signing of the unemployment insurance extension
in late 1991.
Structural Unemployment
One way to think about "structural" unemployment: real wages get in a
sense "stuck" too high.
Insiders and outsiders: unions that respond to desires of employed
workers (and not of unemployed ex- or future members) alone. But hard
to see how this should lead to overall unemployment (rather
than just unemployment in a particular sector or industry). (Unions
as source of collective voice: communication channels; exit vs.
voice; Freeman and Medoff--unionization raises private-sector
productivity by 15%, but wages by 25% or so... Hence rational for
firm to want to union-bust, but far from good for social
welfare).
"Fairness" as a factor keeping real wages high. Employers' (and
employees') beliefs that further wage reductions would be "unfair".
- Digression on the minimum wage: a way to transfer purchasing
power to low-wage workers at the price of increasing unemployment
among low-wage workers.
- Should we worry about this increase in unemployment (tenure
is short already, so a lot of mixing of beneficiaries and
losers; but do we want to further accelerate this trend?)
- Not much sign of large effects on unemployment.
- But also a lot of minimum-wage workers are in relatively
high-income households.
As a result, I tend to be an EITC person (with all its flaws).
Efficiency wage models. The starting point here is
that--whether because of fairness, or fear of losing a good thing, or
whatever--that workers' productivity is in general positively related
to the wages that workers are paid. Thus firms might well not
wish to cut wages--even when there are people outside the business
clamoring for jobs--because they fear the adverse consequences for
worker effort and productivity.
Suppose that a worker's productivity depends on E, on work
effort. Think about a firm trying to maximize its profits with a
very over simplified production function:
Profit = k(EL) - wL
Take the derivative of profit with respect to the wage, noting that
effort depends on the wage as well:
d(Profit)/dw = k[L(dE/dw)] - L
Setting the change in profit equal to zero:
k[dE/dw] = 1
Now take the derivative of profit with respect to the number of
workers:
Profit = k(EL) - wL
d(Profit/dL) = kE - w; setting the change in profit to zero means: k
= w/E
So: [w/E][dE//dw] = 1...
[dE/E][dw/w] = 1
So the business is going to choose the wage such that the
elasticity of effort with respect to the wage is equal to 1:
it will keep on raising wages as long as a 1% rise in wages brings
forth a 1% or greater rise in effort whether or not there are
unemployed workers out there.
This is a rather depressing line of thought: it seems to suggest that
the rate of unemployment could go anywhere at all. Things probably
aren't that bad. A higher level of unemployment probably makes effort
more responsive to the wage: when jobs are hard to find, you care
more about making employers happy in order to keep a good one.
Reserve army of the unemployed.
Henry Ford and the $5 day; $2 a day as the average wage for
semi-skilled workers in 1913 in Detroit. Turnover at Ford more than
370 percent per year. Turnover rate down to 20% in 1915.
Cyclical Unemployment
Cyclical unemployment we have already talked about a lot under the
headings of aggregate supply and
the "Phillips curve".
Implication of Phillips curve: cyclical unemployment more-or-less
balances out. You want to reduce unemployment, figure out how to
reduce structural unemployment:
- Job search assistance
- Ways to diminish "efficiency wage" problems
>
